Autism spectrum condition (ASC) or autism spectrum disorder (ASD) is primarily identified with the help of behavioral indications encompassing social, sensory and motor characteristics. Although categorized, recurring motor actions are measured during diagnosis, quantifiable measures that ascertain kinematic physiognomies in the movement configurations of autistic persons are not adequately studied, hindering the advances in understanding the etiology of motor mutilation. Subject aspects such as behavioral characters that influences ASD need further exploration. Presently, limited autism datasets concomitant with screening ASD are available, and a majority of them are genetic. Hence, in this study, we used a dataset related to autism screening enveloping ten behavioral and ten personal attributes that have been effective in diagnosing ASD cases from controls in behavior science. ASD diagnosis is time exhaustive and uneconomical. The burgeoning ASD cases worldwide mandate a need for the fast and economical screening tool. Our study aimed to implement an artificial neural network with the Levenberg- Marquardt algorithm to detect ASD and examine its predictive accuracy. Consecutively, develop a clinical decision support system for early ASD identification.
IRJET- Plant Disease Detection and Classification using Image Processing a...IRJET Journal
This document describes a method for detecting and classifying plant diseases using image processing and artificial neural networks. The method involves preprocessing images through grayscaling, resizing and filtering. K-means clustering is used to segment infected leaf regions. Features are extracted from segmented images and fed into feedforward and cascaded feedforward neural networks for disease classification. The method achieved accurate classification of several common plant diseases with fewer iterations and better performance than traditional feedforward backpropagation neural networks. This automatic disease detection approach could help improve agricultural productivity by facilitating early detection on large farms.
Accuracy, Sensitivity and Specificity Measurement of Various Classification T...IOSR Journals
This document compares the accuracy, sensitivity, and specificity of various classification techniques when applied to healthcare data on diabetes. It analyzes several algorithms implemented in Weka (Multilayer Perception, Bayes Network, J48graft, JRip) and other tools (PNN, LVQ, FFN, etc. in MATLAB and GINI in RapidMiner) on a diabetes dataset. The results show that J48graft had the highest accuracy at 81.33% while PNN had the highest sensitivity at 63.33% and DTDN had the highest specificity at 88.8% based on calculations using true/false positive/negative values. Therefore, different algorithms performed best for different evaluation metrics on this healthcare
This document discusses using data mining and neural networks to identify negatively influenced factors in patients with liver disorders. It presents a neural network model with liver enzyme values as inputs and physical/biological symptoms as hidden nodes to classify patients as having alcoholic fatty liver disorder. The network was trained using backpropagation to minimize error. Analysis of variance was used to identify relationships between input and hidden nodes. Negatively weighted hidden nodes were analyzed to determine influential epidemiological factors for liver disorder patients.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
Classification of physiological diseases using eeg signals and machine learni...eSAT Journals
Abstract
In rural areas providing advanced diagnostics for various health disorders is not possible in countries like India. With latest technological breakthrough, brain signals (EEG signal) capturing devices are available at rate less 50$. If these brain signals can be used to predict any Physiological disorders like heart problem, kidney problems etc., then these EEG devices can be provided to rural health care centre for preliminary investigation and on diagnosis the patient can move to city hospitals for diagnostics and treatment. In this project, we provide a solution of identifying physiological problems using EEG signals and use machine learning techniques for diagnosis.
Keywords: EEG Signals, EEG Frame, Feature Extraction
Classification of medical datasets using back propagation neural network powe...IJECEIAES
The classification is a one of the most indispensable domains in the data mining and machine learning. The classification process has a good reputation in the area of diseases diagnosis by computer systems where the progress in smart technologies of computer can be invested in diagnosing various diseases based on data of real patients documented in databases. The paper introduced a methodology for diagnosing a set of diseases including two types of cancer (breast cancer and lung), two datasets for diabetes and heart attack. Back Propagation Neural Network plays the role of classifier. The performance of neural net is enhanced by using the genetic algorithm which provides the classifier with the optimal features to raise the classification rate to the highest possible. The system showed high efficiency in dealing with databases differs from each other in size, number of features and nature of the data and this is what the results illustrated, where the ratio of the classification reached to 100% in most datasets).
A clonal based algorithm for the reconstruction of genetic network using s sy...eSAT Journals
Abstract Motivation: Gene regulatory network is the network based approach to represent the interactions between genes. DNA microarray is the most widely used technology for extracting the relationships between thousands of genes simultaneously. Gene microarray experiment provides the gene expression data for a particular condition and varying time periods. The expression of a particular gene depends upon the biological conditions and other genes. In this paper, we propose a new method for the analysis of microarray data. The proposed method makes use of S-system, which is a well-accepted model for the gene regulatory network reconstruction. Since the problem has multiple solutions, we have to identify an optimized solution. Evolutionary algorithms have been used to solve such problems. Though there are a number of attempts already been carried out by various researchers, the solutions are still not that satisfactory with respect to the time taken and the degree of accuracy achieved. Therefore, there is a need of huge amount further work in this topic for achieving solutions with improved performances. Results: In this work, we have proposed Clonal selection algorithm for identifying optimal gene regulatory network. The approach is tested on the real life data: SOS Ecoli DNA repairing gene expression data. It is observed that the proposed algorithm converges much faster and provides better results than the existing algorithms. Index Terms: Microarray analysis, Evolutionary Algorithm, Artificial Immune System, S-system, Gene Regulatory Network, SOS Ecoli DNA repairing, Clonal Selection Algorithm.
Application of Hybrid Genetic Algorithm Using Artificial Neural Network in Da...IOSRjournaljce
The main purpose of data mining is to extract knowledge from large amount of data. Artificial Neural network (ANN) has already been applied in a variety of domains with remarkable success. This paper presents the application of hybrid model for stroke disease that integrates Genetic algorithm and back propagation algorithm. Selecting a good subset of features, without sacrificing accuracy, is of great importance for neural networks to be successfully applied to the area. In addition the hybrid model that leads to further improvised categorization, accuracy compared to the result produced by genetic algorithm alone. In this study, a new hybrid model of Neural Networks and Genetic Algorithm (GA) to initialize and optimize the connection weights of ANN so as to improve the performance of the ANN and the same has been applied in a medical problem of predicting stroke disease for verification of the results.
IRJET- Plant Disease Detection and Classification using Image Processing a...IRJET Journal
This document describes a method for detecting and classifying plant diseases using image processing and artificial neural networks. The method involves preprocessing images through grayscaling, resizing and filtering. K-means clustering is used to segment infected leaf regions. Features are extracted from segmented images and fed into feedforward and cascaded feedforward neural networks for disease classification. The method achieved accurate classification of several common plant diseases with fewer iterations and better performance than traditional feedforward backpropagation neural networks. This automatic disease detection approach could help improve agricultural productivity by facilitating early detection on large farms.
Accuracy, Sensitivity and Specificity Measurement of Various Classification T...IOSR Journals
This document compares the accuracy, sensitivity, and specificity of various classification techniques when applied to healthcare data on diabetes. It analyzes several algorithms implemented in Weka (Multilayer Perception, Bayes Network, J48graft, JRip) and other tools (PNN, LVQ, FFN, etc. in MATLAB and GINI in RapidMiner) on a diabetes dataset. The results show that J48graft had the highest accuracy at 81.33% while PNN had the highest sensitivity at 63.33% and DTDN had the highest specificity at 88.8% based on calculations using true/false positive/negative values. Therefore, different algorithms performed best for different evaluation metrics on this healthcare
This document discusses using data mining and neural networks to identify negatively influenced factors in patients with liver disorders. It presents a neural network model with liver enzyme values as inputs and physical/biological symptoms as hidden nodes to classify patients as having alcoholic fatty liver disorder. The network was trained using backpropagation to minimize error. Analysis of variance was used to identify relationships between input and hidden nodes. Negatively weighted hidden nodes were analyzed to determine influential epidemiological factors for liver disorder patients.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
Classification of physiological diseases using eeg signals and machine learni...eSAT Journals
Abstract
In rural areas providing advanced diagnostics for various health disorders is not possible in countries like India. With latest technological breakthrough, brain signals (EEG signal) capturing devices are available at rate less 50$. If these brain signals can be used to predict any Physiological disorders like heart problem, kidney problems etc., then these EEG devices can be provided to rural health care centre for preliminary investigation and on diagnosis the patient can move to city hospitals for diagnostics and treatment. In this project, we provide a solution of identifying physiological problems using EEG signals and use machine learning techniques for diagnosis.
Keywords: EEG Signals, EEG Frame, Feature Extraction
Classification of medical datasets using back propagation neural network powe...IJECEIAES
The classification is a one of the most indispensable domains in the data mining and machine learning. The classification process has a good reputation in the area of diseases diagnosis by computer systems where the progress in smart technologies of computer can be invested in diagnosing various diseases based on data of real patients documented in databases. The paper introduced a methodology for diagnosing a set of diseases including two types of cancer (breast cancer and lung), two datasets for diabetes and heart attack. Back Propagation Neural Network plays the role of classifier. The performance of neural net is enhanced by using the genetic algorithm which provides the classifier with the optimal features to raise the classification rate to the highest possible. The system showed high efficiency in dealing with databases differs from each other in size, number of features and nature of the data and this is what the results illustrated, where the ratio of the classification reached to 100% in most datasets).
A clonal based algorithm for the reconstruction of genetic network using s sy...eSAT Journals
Abstract Motivation: Gene regulatory network is the network based approach to represent the interactions between genes. DNA microarray is the most widely used technology for extracting the relationships between thousands of genes simultaneously. Gene microarray experiment provides the gene expression data for a particular condition and varying time periods. The expression of a particular gene depends upon the biological conditions and other genes. In this paper, we propose a new method for the analysis of microarray data. The proposed method makes use of S-system, which is a well-accepted model for the gene regulatory network reconstruction. Since the problem has multiple solutions, we have to identify an optimized solution. Evolutionary algorithms have been used to solve such problems. Though there are a number of attempts already been carried out by various researchers, the solutions are still not that satisfactory with respect to the time taken and the degree of accuracy achieved. Therefore, there is a need of huge amount further work in this topic for achieving solutions with improved performances. Results: In this work, we have proposed Clonal selection algorithm for identifying optimal gene regulatory network. The approach is tested on the real life data: SOS Ecoli DNA repairing gene expression data. It is observed that the proposed algorithm converges much faster and provides better results than the existing algorithms. Index Terms: Microarray analysis, Evolutionary Algorithm, Artificial Immune System, S-system, Gene Regulatory Network, SOS Ecoli DNA repairing, Clonal Selection Algorithm.
Application of Hybrid Genetic Algorithm Using Artificial Neural Network in Da...IOSRjournaljce
The main purpose of data mining is to extract knowledge from large amount of data. Artificial Neural network (ANN) has already been applied in a variety of domains with remarkable success. This paper presents the application of hybrid model for stroke disease that integrates Genetic algorithm and back propagation algorithm. Selecting a good subset of features, without sacrificing accuracy, is of great importance for neural networks to be successfully applied to the area. In addition the hybrid model that leads to further improvised categorization, accuracy compared to the result produced by genetic algorithm alone. In this study, a new hybrid model of Neural Networks and Genetic Algorithm (GA) to initialize and optimize the connection weights of ANN so as to improve the performance of the ANN and the same has been applied in a medical problem of predicting stroke disease for verification of the results.
An Artificial Neural Network Model for Neonatal Disease DiagnosisWaqas Tariq
The significance of disease diagnosis by artificial intelligence is not obscure now days. The increasing demand of Artificial Neural Network application for predicting the disease shows better performance in the field of medical decision making. This paper represents the use of artificial neural networks in predicting neonatal disease diagnosis. The proposed technique involves training a Multi Layer Perceptron with a BP learning algorithm to recognize a pattern for the diagnosing and prediction of neonatal diseases. A comparative study of using different training algorithm of MLP, Quick Propagation, Conjugate Gradient Descent, shows the higher prediction accuracy. The Backpropogation algorithm was used to train the ANN architecture and the same has been tested for the various categories of neonatal disease. About 94 cases of different sign and symptoms parameter have been tested in this model. This study exhibits ANN based prediction of neonatal disease and improves the diagnosis accuracy of 75% with higher stability. Key words: Artificial Intelligence, Multi Layer Perceptron, Neural Network, Neonate
IRJET- Prediction of Heart Disease using RNN AlgorithmIRJET Journal
This document discusses using a recurrent neural network (RNN) algorithm to predict heart disease. It proposes a method called prognosis prediction using RNN (PP-RNN) that uses multiple RNNs to learn from patient diagnosis code sequences in order to predict high-risk diseases. The experimental results show that the proposed PP-RNN method can achieve more accurate results than existing methods for predicting heart disease risk. It also provides background on related works using other techniques like decision trees, clustering, and AdaBoost for heart disease prediction.
Delineation of techniques to implement on the enhanced proposed model using d...ijdms
In post genomic era with the advent of new technologies a huge amount of complex molecular data are
generated with high throughput. The management of this biological data is definitely a challenging task
due to complexity and heterogeneity of data for discovering new knowledge. Issues like managing noisy
and incomplete data are needed to be dealt with. Use of data mining in biological domain has made its
inventory success. Discovering new knowledge from the biological data is a major challenge in data
mining technique. The novelty of the proposed model is its combined use of intelligent techniques to classify
the protein sequence faster and efficiently. Use of FFT, fuzzy classifier, String weighted algorithm, gram
encoding method, neural network model and rough set classifier in a single model and in an appropriate
place can enhance the quality of the classification system .Thus the primary challenge is to identify and
classify the large protein sequences in a very fast and easy but intellectual way to decrease the time
complexity and space complexity.
MLTDD : USE OF MACHINE LEARNING TECHNIQUES FOR DIAGNOSIS OF THYROID GLAND DIS...cscpconf
Machine learning algorithms are used to diagnosis for many diseases after very important improvements of classification algorithms as well as having large data sets and high performing computational units. All of these increased the accuracy of these methods. The diagnosis of thyroid gland disorders is one of the application for important classification problem. This study majorly focuses on thyroid gland medical diseases caused by underactive or overactive thyroid glands. The dataset used for the study was taken from UCI repository. Classification of this thyroid disease dataset was a considerable task using decision tree algorithm. The overall
prediction accuracy is 100% for training and in range between 98.7% and 99.8% for testing. In this study, we developed the Machine Learning tool for Thyroid Disease Diagnosis (MLTDD), an Intelligent thyroid gland disease prediction tool in Python, which can effectively help to make the right decision, has been designed using PyDev, which is python IDE for Eclipse.
An approach of re-organizing input dataset to enhance the quality of emotion ...journalBEEI
The purpose of this paper is to propose an approach of re-organizing input data to recognize emotion based on short signal segments and increase the quality of emotional recognition using physiological signals. MIT's long physiological signal set was divided into two new datasets, with shorter and overlapped segments. Three different classification methods (support vector machine, random forest, and multilayer perceptron) were implemented to identify eight emotional states based on statistical features of each segment in these two datasets. By re-organizing the input dataset, the quality of recognition results was enhanced. The random forest shows the best classification result among three implemented classification methods, with an accuracy of 97.72% for eight emotional states, on the overlapped dataset. This approach shows that, by re-organizing the input dataset, the high accuracy of recognition results can be achieved without the use of EEG and ECG signals.
Prediction of Dengue, Diabetes and Swine Flu using Random Forest Classificati...IRJET Journal
This document describes a disease prediction system that uses the Random Forest classification algorithm to predict Dengue, diabetes, and swine flu. The system trains on labeled datasets for each disease. It then takes user-entered symptoms as input and predicts the likelihood of each disease. If a disease is predicted to be positive, the system recommends a specialized doctor. The document discusses related work on disease prediction using data mining techniques. It provides an overview of how the Random Forest algorithm works for classification problems and ensemble learning. The proposed system aims to help users predict diseases and find appropriate doctors for treatment.
The document compares machine learning techniques for identifying fish disease, specifically Epizootic Ulcerative Syndrome (EUS). It evaluates different combinations of feature extraction (HOG, FAST), dimensionality reduction (PCA), and classification (KNN, Neural Network). The proposed combination of FAST feature extraction, PCA dimensionality reduction, and a Neural Network classifier achieved the highest accuracy of 96.3% for identifying EUS-infected fish, outperforming the other combinations tested.
AN EFFICIENT PSO BASED ENSEMBLE CLASSIFICATION MODEL ON HIGH DIMENSIONAL DATA...ijsc
As the size of the biomedical databases are growing day by day, finding an essential features in the disease prediction have become more complex due to high dimensionality and sparsity problems. Also, due to the
availability of a large number of micro-array datasets in the biomedical repositories, it is difficult to analyze, predict and interpret the feature information using the traditional feature selection based classification models. Most of the traditional feature selection based classification algorithms have computational issues such as dimension reduction, uncertainty and class imbalance on microarray datasets. Ensemble classifier is one of the scalable models for extreme learning machine due to its high efficiency, the fast processing speed for real-time applications. The main objective of the feature selection
based ensemble learning models is to classify the high dimensional data with high computational efficiency
and high true positive rate on high dimensional datasets. In this proposed model an optimized Particle swarm optimization (PSO) based Ensemble classification model was developed on high dimensional microarray
datasets. Experimental results proved that the proposed model has high computational efficiency compared to the traditional feature selection based classification models in terms of accuracy , true positive rate and error rate are concerned.
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...Tarun Kumar
In this computing era, image processing has
spread its wings in human life upto the extent that image
has become an integral part of their life. There are various
applications of image processing in the field of commerce,
engineering, graphic design, journalism, architecture and
historical research. In this research work, Image
processing is considered for the analysis of plant leaf
diseases. Plant leaf diseases can be detected based on the
disease symptoms. Here, dataset of disease affected leaves
is considered for experimentation. This dataset contains
the plant leaves suffered from the
AlternariaAlternata,Cercospora Leaf Spot, Anthracnose
andBacterial Blight along with some healthy leaf images.
For this analysis, an autonomous approach of modified
SVM-CS is introduces. Here, concept of cuckoo search is
considered to optimize the classification parameters. These
parameters further help to find more accurate solutions.
This autonomous approach also extracts the healthy
portion and disease affected leaf portion along with the
accuracy of results.
IRJET- Facial Expression Recognition System using Neural Network based on...IRJET Journal
This document describes a facial expression recognition system using a neural network approach. It uses the Japanese Female Facial Expressions (JAFFE) database to classify 7 facial expressions. The system extracts features using 2D discrete cosine transform (DCT), local binary patterns (LBP), and histogram of oriented gradients (HOG). These features are used to create a hybrid feature vector for each image. A single hidden layer feedforward neural network is trained on the feature vectors using different learning algorithms to classify the expressions. Experimental results show that a neural network trained with gradient descent and adaptive learning rate achieves the highest average accuracy of 97.2% for classifying expressions in the JAFFE database.
The Evaluated Measurement of a Combined Genetic Algorithm and Artificial Immu...IJECEIAES
This paper demonstrates a hybrid between two optimization methods which are the Artificial Immune System (AIS) and Genetic Algorithm (GA). The novel algorithm called the immune genetic algorithm (IGA), provides improvement to the results that enable GA and AIS to work separately which is the main objective of this hybrid. Negative selection which is one of the techniques in the AIS, was employed to determine the input variables (populations) of the system. In order to illustrate the effectiveness of the IGA, the comparison with a steady-state GA, AIS, and PSO were also investigated. The testing of the performance was conducted by mathematical testing, problems were divided into single and multiple objectives. The five single objectives were then used to test the modified algorithm, the results showed that IGA performed better than all of the other methods. The DTLZ multi-objective testing functions were then used. The result also illustrated that the modified approach still had the best performance.
IRJET - A Review on Identification and Disease Detection in Plants using Mach...IRJET Journal
This document reviews machine learning techniques for identifying and detecting plant diseases. It discusses how techniques like artificial neural networks, support vector machines, K-nearest neighbors classification and fuzzy c-means clustering have been applied to identify diseases in crops like rice, potatoes, cucumbers and grapes. The techniques analyze images of plant leaves to extract features and classify whether the plant has a disease or not. The document also outlines the common stages of disease identification using machine learning, which include preprocessing images, segmentation, feature extraction, classification and disease identification.
ENHANCED BREAST CANCER RECOGNITION BASED ON ROTATION FOREST FEATURE SELECTIO...cscpconf
Optimization problems are dominantly being solved using Computational Intelligence. One of
the issues that can be addressed in this context is problems related to attribute subset selection
evaluation. This paper presents a computational intelligence technique for solving the
optimization problem using a proposed model called Modified Genetic Search Algorithms
(MGSA) that avoids local bad search space with merit and scaled fitness variables, detecting
and deleting bad candidate chromosomes, thereby reducing the number of individual
chromosomes from search space and subsequent iterations in next generations. This paper aims
to show that Rotation forest ensembles are useful in the feature selection method. The base
classifier is multinomial logistic regression method integrated with Haar wavelets as projection
filter and reproducing the ranks of each features with 10 fold cross validation method. It also
discusses the main findings and concludes with promising result of the proposed model. It
explores the combination of MGSA for optimization with Naïve Bayes classification. The result
obtained using proposed model MGSA is validated mathematically using Principal Component
Analysis. The goal is to improve the accuracy and quality of diagnosis of Breast cancer disease
with robust machine learning algorithms. As compared to other works in literature survey,
experimental results achieved in this paper show better results with statistical inferenc
Preprocessing and Classification in WEKA Using Different ClassifiersIJERA Editor
Data mining is a process of extracting information from a dataset and transform it into understandable structure
for further use, also it discovers patterns in large data sets [1]. Data mining has number of important techniques
such as preprocessing, classification. Classification is one such technique which is based on supervised learning.
It is a technique used for predicting group membership for the data instance. Here in this paper we use
preprocessing, classification on diabetes database. Here we apply classifiers on this database and compare the
result based on certain parameters using WEKA. 77.2 million people in India are suffering from pre diabetes.
ICMR estimates that around 65.1million are diabetes patients. Globally in year 2010, 227 to 285 million people
had diabetes, out of that 90% cases are related to type 2 ,this is equal to 3.3% of the population with equal rates
in both women and men in 2011 it resulted in 1.4 million deaths worldwide making it the leading cause of
death.
Identification of Disease in Leaves using Genetic Algorithmijtsrd
Plant disease is an impairment of normal state of a plant that interrupts or modifies its vital functions. Many leaf diseases are caused by pathogens. Agriculture is the mains try of the Indian economy. Perception of human eye is not so much stronger so as to observe minute variation in the infected part of leaf. In this paper, we are providing software solution to automatically detect and classify plant leaf diseases. In this we are using image processing techniques to classify diseases and quickly diagnosis can be carried out as per disease. This approach will enhance productivity of crops. It includes image processing techniques starting from image acquisition, preprocessing, testing, and training. K. Beulah Suganthy ""Identification of Disease in Leaves using Genetic Algorithm"" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-3 | Issue-3 , April 2019, URL: https://www.ijtsrd.com/papers/ijtsrd22901.pdf
Paper URL: https://www.ijtsrd.com/engineering/electronics-and-communication-engineering/22901/identification-of-disease-in-leaves-using-genetic-algorithm/k-beulah-suganthy
Ontheclassificationof ee gsignalbyusingansvmbasedalgorythmKarthik S
Major depression, also termed as major depressive disorder (MDD),
unipolar depression, clinical depression, or even simply depression, is a
mental illness. According to the World Health Organization (WHO),
depression has been identified as a leading cause of functional disability, worldwide. About 300 million people have been reported suffering
from depression, globally.1 In addition to the functional disability caused
by depression, it may lead to suicide ideations. Moreover, the treatment
management for depression has been challenging due to multiple factors,
such as the suitable selection of medication for a patient being based on
the subjective experience of clinicians and which might not be appropriate for the patient and could result into unsuccessful treatment trials.
Another implication is that the patient may stop the treatment.
In this chapter, the topics covered in this book are introduced by providing a basic explanation of the relevant concepts which will be elaborated on in later chapters. More specifically, this chapter explores the
possibilities of utilizing electroencephalogram (EEG) as an objective
method for the diagnosis and treatment efficacy assessment for depression. Also, depression will be discussed from different perspectives
such as its subtypes, signs and symptoms, the challenges associated
with treating depression, an overview of the literature involving EEG
studies for depression, EEG as a modality, and the basics of an EEGbased machine learning (ML) approach
IRJET- Detection and Classification of Leaf DiseasesIRJET Journal
This document presents a method for detecting and classifying leaf diseases using image processing techniques. The method involves 4 main phases: 1) Image preprocessing including noise removal and color space transformation. 2) Image segmentation using k-means clustering to separate healthy and infected tissue. 3) Feature extraction of texture characteristics. 4) Classification of the disease using a support vector machine model. The results diagnose the disease name and percentage of leaf area infected to help farmers quickly identify and respond to plant diseases.
ROLE OF CERTAINTY FACTOR IN GENERATING ROUGH-FUZZY RULEIJCSEA Journal
The generation of effective feature-based rules is essential to the development of any intelligent system. This paper presents an approach that integrates a powerful fuzzy rule generation algorithm with a rough set-assisted feature reduction method to generate diagnostic rule with a certainty factor. Certainty factor of each rule is calculated by considering both the membership value of each linguistic term introduced at time of fuzzyfication of data as well as possibility values, due to inconsistent data, generated by rough set theory at time of rule generation. In time of knowledge inferencing in an intelligent system, certainty factor of each rule will play an important role to find out the appropriate rule to be selected. Experimental results demonstrate the superiority of our approach.
This document describes two machine learning techniques, particle swarm optimization with support vector machines (PSO-SVM) and recursive feature elimination with support vector machines (RFE-SVM), that were used to classify autism neuroimaging data from the Autism Brain Imaging Data Exchange database. PSO-SVM was used to select discriminative features for classification, while RFE-SVM ranked features by importance. Both techniques aimed to improve classification accuracy and reduce overfitting by selecting optimal feature subsets from the high-dimensional neuroimaging data. The results could help develop brain-based diagnostic criteria for autism.
MULTI-PARAMETER BASED PERFORMANCE EVALUATION OF CLASSIFICATION ALGORITHMSijcsit
Diabetes disease is amongst the most common disease in India. It affects patient’s health and also leads to
other chronic diseases. Prediction of diabetes plays a significant role in saving of life and cost. Predicting
diabetes in human body is a challenging task because it depends on several factors. Few studies have reported the performance of classification algorithms in terms of accuracy. Results in these studies are difficult and complex to understand by medical practitioner and also lack in terms of visual aids as they arepresented in pure text format. This reported survey uses ROC and PRC graphical measures toimproveunderstanding of results. A detailed parameter wise discussion of comparison is also presented which lacksin other reported surveys. Execution time, Accuracy, TP Rate, FP Rate, Precision, Recall, F Measureparameters are used for comparative analysis and Confusion Matrix is prepared for quick review of each
algorithm. Ten fold cross validation method is used for estimation of prediction model. Different sets of
classification algorithms are analyzed on diabetes dataset acquired from UCI repository
An Efficient PSO Based Ensemble Classification Model on High Dimensional Data...ijsc
This summary provides the high-level information from the document in 3 sentences:
The document proposes a Particle Swarm Optimization (PSO) based ensemble classification model to improve classification of high-dimensional biomedical datasets. It develops an optimized PSO technique to select optimal features and initialize weights for base classifiers in the ensemble model. Experimental results on microarray datasets show the proposed model achieves higher accuracy, true positive rate, and lower error rate compared to traditional feature selection based classification models.
IRJET- Comparative Study of Machine Learning Models for Alzheimer’s Detec...IRJET Journal
This document presents a comparative study of machine learning models for detecting Alzheimer's disease. The study uses MRI data to extract features like brain volume and uses those as inputs to various machine learning models like logistic regression, support vector machines, decision trees, random forests and AdaBoost. The performance of each model is evaluated using metrics like accuracy, sensitivity and specificity. The results show that the random forest model performs the best with the highest prediction rates, indicating it has potential for accurate early detection of Alzheimer's using MRI data.
An Artificial Neural Network Model for Neonatal Disease DiagnosisWaqas Tariq
The significance of disease diagnosis by artificial intelligence is not obscure now days. The increasing demand of Artificial Neural Network application for predicting the disease shows better performance in the field of medical decision making. This paper represents the use of artificial neural networks in predicting neonatal disease diagnosis. The proposed technique involves training a Multi Layer Perceptron with a BP learning algorithm to recognize a pattern for the diagnosing and prediction of neonatal diseases. A comparative study of using different training algorithm of MLP, Quick Propagation, Conjugate Gradient Descent, shows the higher prediction accuracy. The Backpropogation algorithm was used to train the ANN architecture and the same has been tested for the various categories of neonatal disease. About 94 cases of different sign and symptoms parameter have been tested in this model. This study exhibits ANN based prediction of neonatal disease and improves the diagnosis accuracy of 75% with higher stability. Key words: Artificial Intelligence, Multi Layer Perceptron, Neural Network, Neonate
IRJET- Prediction of Heart Disease using RNN AlgorithmIRJET Journal
This document discusses using a recurrent neural network (RNN) algorithm to predict heart disease. It proposes a method called prognosis prediction using RNN (PP-RNN) that uses multiple RNNs to learn from patient diagnosis code sequences in order to predict high-risk diseases. The experimental results show that the proposed PP-RNN method can achieve more accurate results than existing methods for predicting heart disease risk. It also provides background on related works using other techniques like decision trees, clustering, and AdaBoost for heart disease prediction.
Delineation of techniques to implement on the enhanced proposed model using d...ijdms
In post genomic era with the advent of new technologies a huge amount of complex molecular data are
generated with high throughput. The management of this biological data is definitely a challenging task
due to complexity and heterogeneity of data for discovering new knowledge. Issues like managing noisy
and incomplete data are needed to be dealt with. Use of data mining in biological domain has made its
inventory success. Discovering new knowledge from the biological data is a major challenge in data
mining technique. The novelty of the proposed model is its combined use of intelligent techniques to classify
the protein sequence faster and efficiently. Use of FFT, fuzzy classifier, String weighted algorithm, gram
encoding method, neural network model and rough set classifier in a single model and in an appropriate
place can enhance the quality of the classification system .Thus the primary challenge is to identify and
classify the large protein sequences in a very fast and easy but intellectual way to decrease the time
complexity and space complexity.
MLTDD : USE OF MACHINE LEARNING TECHNIQUES FOR DIAGNOSIS OF THYROID GLAND DIS...cscpconf
Machine learning algorithms are used to diagnosis for many diseases after very important improvements of classification algorithms as well as having large data sets and high performing computational units. All of these increased the accuracy of these methods. The diagnosis of thyroid gland disorders is one of the application for important classification problem. This study majorly focuses on thyroid gland medical diseases caused by underactive or overactive thyroid glands. The dataset used for the study was taken from UCI repository. Classification of this thyroid disease dataset was a considerable task using decision tree algorithm. The overall
prediction accuracy is 100% for training and in range between 98.7% and 99.8% for testing. In this study, we developed the Machine Learning tool for Thyroid Disease Diagnosis (MLTDD), an Intelligent thyroid gland disease prediction tool in Python, which can effectively help to make the right decision, has been designed using PyDev, which is python IDE for Eclipse.
An approach of re-organizing input dataset to enhance the quality of emotion ...journalBEEI
The purpose of this paper is to propose an approach of re-organizing input data to recognize emotion based on short signal segments and increase the quality of emotional recognition using physiological signals. MIT's long physiological signal set was divided into two new datasets, with shorter and overlapped segments. Three different classification methods (support vector machine, random forest, and multilayer perceptron) were implemented to identify eight emotional states based on statistical features of each segment in these two datasets. By re-organizing the input dataset, the quality of recognition results was enhanced. The random forest shows the best classification result among three implemented classification methods, with an accuracy of 97.72% for eight emotional states, on the overlapped dataset. This approach shows that, by re-organizing the input dataset, the high accuracy of recognition results can be achieved without the use of EEG and ECG signals.
Prediction of Dengue, Diabetes and Swine Flu using Random Forest Classificati...IRJET Journal
This document describes a disease prediction system that uses the Random Forest classification algorithm to predict Dengue, diabetes, and swine flu. The system trains on labeled datasets for each disease. It then takes user-entered symptoms as input and predicts the likelihood of each disease. If a disease is predicted to be positive, the system recommends a specialized doctor. The document discusses related work on disease prediction using data mining techniques. It provides an overview of how the Random Forest algorithm works for classification problems and ensemble learning. The proposed system aims to help users predict diseases and find appropriate doctors for treatment.
The document compares machine learning techniques for identifying fish disease, specifically Epizootic Ulcerative Syndrome (EUS). It evaluates different combinations of feature extraction (HOG, FAST), dimensionality reduction (PCA), and classification (KNN, Neural Network). The proposed combination of FAST feature extraction, PCA dimensionality reduction, and a Neural Network classifier achieved the highest accuracy of 96.3% for identifying EUS-infected fish, outperforming the other combinations tested.
AN EFFICIENT PSO BASED ENSEMBLE CLASSIFICATION MODEL ON HIGH DIMENSIONAL DATA...ijsc
As the size of the biomedical databases are growing day by day, finding an essential features in the disease prediction have become more complex due to high dimensionality and sparsity problems. Also, due to the
availability of a large number of micro-array datasets in the biomedical repositories, it is difficult to analyze, predict and interpret the feature information using the traditional feature selection based classification models. Most of the traditional feature selection based classification algorithms have computational issues such as dimension reduction, uncertainty and class imbalance on microarray datasets. Ensemble classifier is one of the scalable models for extreme learning machine due to its high efficiency, the fast processing speed for real-time applications. The main objective of the feature selection
based ensemble learning models is to classify the high dimensional data with high computational efficiency
and high true positive rate on high dimensional datasets. In this proposed model an optimized Particle swarm optimization (PSO) based Ensemble classification model was developed on high dimensional microarray
datasets. Experimental results proved that the proposed model has high computational efficiency compared to the traditional feature selection based classification models in terms of accuracy , true positive rate and error rate are concerned.
Plant Leaf Disease Analysis using Image Processing Technique with Modified SV...Tarun Kumar
In this computing era, image processing has
spread its wings in human life upto the extent that image
has become an integral part of their life. There are various
applications of image processing in the field of commerce,
engineering, graphic design, journalism, architecture and
historical research. In this research work, Image
processing is considered for the analysis of plant leaf
diseases. Plant leaf diseases can be detected based on the
disease symptoms. Here, dataset of disease affected leaves
is considered for experimentation. This dataset contains
the plant leaves suffered from the
AlternariaAlternata,Cercospora Leaf Spot, Anthracnose
andBacterial Blight along with some healthy leaf images.
For this analysis, an autonomous approach of modified
SVM-CS is introduces. Here, concept of cuckoo search is
considered to optimize the classification parameters. These
parameters further help to find more accurate solutions.
This autonomous approach also extracts the healthy
portion and disease affected leaf portion along with the
accuracy of results.
IRJET- Facial Expression Recognition System using Neural Network based on...IRJET Journal
This document describes a facial expression recognition system using a neural network approach. It uses the Japanese Female Facial Expressions (JAFFE) database to classify 7 facial expressions. The system extracts features using 2D discrete cosine transform (DCT), local binary patterns (LBP), and histogram of oriented gradients (HOG). These features are used to create a hybrid feature vector for each image. A single hidden layer feedforward neural network is trained on the feature vectors using different learning algorithms to classify the expressions. Experimental results show that a neural network trained with gradient descent and adaptive learning rate achieves the highest average accuracy of 97.2% for classifying expressions in the JAFFE database.
The Evaluated Measurement of a Combined Genetic Algorithm and Artificial Immu...IJECEIAES
This paper demonstrates a hybrid between two optimization methods which are the Artificial Immune System (AIS) and Genetic Algorithm (GA). The novel algorithm called the immune genetic algorithm (IGA), provides improvement to the results that enable GA and AIS to work separately which is the main objective of this hybrid. Negative selection which is one of the techniques in the AIS, was employed to determine the input variables (populations) of the system. In order to illustrate the effectiveness of the IGA, the comparison with a steady-state GA, AIS, and PSO were also investigated. The testing of the performance was conducted by mathematical testing, problems were divided into single and multiple objectives. The five single objectives were then used to test the modified algorithm, the results showed that IGA performed better than all of the other methods. The DTLZ multi-objective testing functions were then used. The result also illustrated that the modified approach still had the best performance.
IRJET - A Review on Identification and Disease Detection in Plants using Mach...IRJET Journal
This document reviews machine learning techniques for identifying and detecting plant diseases. It discusses how techniques like artificial neural networks, support vector machines, K-nearest neighbors classification and fuzzy c-means clustering have been applied to identify diseases in crops like rice, potatoes, cucumbers and grapes. The techniques analyze images of plant leaves to extract features and classify whether the plant has a disease or not. The document also outlines the common stages of disease identification using machine learning, which include preprocessing images, segmentation, feature extraction, classification and disease identification.
ENHANCED BREAST CANCER RECOGNITION BASED ON ROTATION FOREST FEATURE SELECTIO...cscpconf
Optimization problems are dominantly being solved using Computational Intelligence. One of
the issues that can be addressed in this context is problems related to attribute subset selection
evaluation. This paper presents a computational intelligence technique for solving the
optimization problem using a proposed model called Modified Genetic Search Algorithms
(MGSA) that avoids local bad search space with merit and scaled fitness variables, detecting
and deleting bad candidate chromosomes, thereby reducing the number of individual
chromosomes from search space and subsequent iterations in next generations. This paper aims
to show that Rotation forest ensembles are useful in the feature selection method. The base
classifier is multinomial logistic regression method integrated with Haar wavelets as projection
filter and reproducing the ranks of each features with 10 fold cross validation method. It also
discusses the main findings and concludes with promising result of the proposed model. It
explores the combination of MGSA for optimization with Naïve Bayes classification. The result
obtained using proposed model MGSA is validated mathematically using Principal Component
Analysis. The goal is to improve the accuracy and quality of diagnosis of Breast cancer disease
with robust machine learning algorithms. As compared to other works in literature survey,
experimental results achieved in this paper show better results with statistical inferenc
Preprocessing and Classification in WEKA Using Different ClassifiersIJERA Editor
Data mining is a process of extracting information from a dataset and transform it into understandable structure
for further use, also it discovers patterns in large data sets [1]. Data mining has number of important techniques
such as preprocessing, classification. Classification is one such technique which is based on supervised learning.
It is a technique used for predicting group membership for the data instance. Here in this paper we use
preprocessing, classification on diabetes database. Here we apply classifiers on this database and compare the
result based on certain parameters using WEKA. 77.2 million people in India are suffering from pre diabetes.
ICMR estimates that around 65.1million are diabetes patients. Globally in year 2010, 227 to 285 million people
had diabetes, out of that 90% cases are related to type 2 ,this is equal to 3.3% of the population with equal rates
in both women and men in 2011 it resulted in 1.4 million deaths worldwide making it the leading cause of
death.
Identification of Disease in Leaves using Genetic Algorithmijtsrd
Plant disease is an impairment of normal state of a plant that interrupts or modifies its vital functions. Many leaf diseases are caused by pathogens. Agriculture is the mains try of the Indian economy. Perception of human eye is not so much stronger so as to observe minute variation in the infected part of leaf. In this paper, we are providing software solution to automatically detect and classify plant leaf diseases. In this we are using image processing techniques to classify diseases and quickly diagnosis can be carried out as per disease. This approach will enhance productivity of crops. It includes image processing techniques starting from image acquisition, preprocessing, testing, and training. K. Beulah Suganthy ""Identification of Disease in Leaves using Genetic Algorithm"" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-3 | Issue-3 , April 2019, URL: https://www.ijtsrd.com/papers/ijtsrd22901.pdf
Paper URL: https://www.ijtsrd.com/engineering/electronics-and-communication-engineering/22901/identification-of-disease-in-leaves-using-genetic-algorithm/k-beulah-suganthy
Ontheclassificationof ee gsignalbyusingansvmbasedalgorythmKarthik S
Major depression, also termed as major depressive disorder (MDD),
unipolar depression, clinical depression, or even simply depression, is a
mental illness. According to the World Health Organization (WHO),
depression has been identified as a leading cause of functional disability, worldwide. About 300 million people have been reported suffering
from depression, globally.1 In addition to the functional disability caused
by depression, it may lead to suicide ideations. Moreover, the treatment
management for depression has been challenging due to multiple factors,
such as the suitable selection of medication for a patient being based on
the subjective experience of clinicians and which might not be appropriate for the patient and could result into unsuccessful treatment trials.
Another implication is that the patient may stop the treatment.
In this chapter, the topics covered in this book are introduced by providing a basic explanation of the relevant concepts which will be elaborated on in later chapters. More specifically, this chapter explores the
possibilities of utilizing electroencephalogram (EEG) as an objective
method for the diagnosis and treatment efficacy assessment for depression. Also, depression will be discussed from different perspectives
such as its subtypes, signs and symptoms, the challenges associated
with treating depression, an overview of the literature involving EEG
studies for depression, EEG as a modality, and the basics of an EEGbased machine learning (ML) approach
IRJET- Detection and Classification of Leaf DiseasesIRJET Journal
This document presents a method for detecting and classifying leaf diseases using image processing techniques. The method involves 4 main phases: 1) Image preprocessing including noise removal and color space transformation. 2) Image segmentation using k-means clustering to separate healthy and infected tissue. 3) Feature extraction of texture characteristics. 4) Classification of the disease using a support vector machine model. The results diagnose the disease name and percentage of leaf area infected to help farmers quickly identify and respond to plant diseases.
ROLE OF CERTAINTY FACTOR IN GENERATING ROUGH-FUZZY RULEIJCSEA Journal
The generation of effective feature-based rules is essential to the development of any intelligent system. This paper presents an approach that integrates a powerful fuzzy rule generation algorithm with a rough set-assisted feature reduction method to generate diagnostic rule with a certainty factor. Certainty factor of each rule is calculated by considering both the membership value of each linguistic term introduced at time of fuzzyfication of data as well as possibility values, due to inconsistent data, generated by rough set theory at time of rule generation. In time of knowledge inferencing in an intelligent system, certainty factor of each rule will play an important role to find out the appropriate rule to be selected. Experimental results demonstrate the superiority of our approach.
This document describes two machine learning techniques, particle swarm optimization with support vector machines (PSO-SVM) and recursive feature elimination with support vector machines (RFE-SVM), that were used to classify autism neuroimaging data from the Autism Brain Imaging Data Exchange database. PSO-SVM was used to select discriminative features for classification, while RFE-SVM ranked features by importance. Both techniques aimed to improve classification accuracy and reduce overfitting by selecting optimal feature subsets from the high-dimensional neuroimaging data. The results could help develop brain-based diagnostic criteria for autism.
MULTI-PARAMETER BASED PERFORMANCE EVALUATION OF CLASSIFICATION ALGORITHMSijcsit
Diabetes disease is amongst the most common disease in India. It affects patient’s health and also leads to
other chronic diseases. Prediction of diabetes plays a significant role in saving of life and cost. Predicting
diabetes in human body is a challenging task because it depends on several factors. Few studies have reported the performance of classification algorithms in terms of accuracy. Results in these studies are difficult and complex to understand by medical practitioner and also lack in terms of visual aids as they arepresented in pure text format. This reported survey uses ROC and PRC graphical measures toimproveunderstanding of results. A detailed parameter wise discussion of comparison is also presented which lacksin other reported surveys. Execution time, Accuracy, TP Rate, FP Rate, Precision, Recall, F Measureparameters are used for comparative analysis and Confusion Matrix is prepared for quick review of each
algorithm. Ten fold cross validation method is used for estimation of prediction model. Different sets of
classification algorithms are analyzed on diabetes dataset acquired from UCI repository
An Efficient PSO Based Ensemble Classification Model on High Dimensional Data...ijsc
This summary provides the high-level information from the document in 3 sentences:
The document proposes a Particle Swarm Optimization (PSO) based ensemble classification model to improve classification of high-dimensional biomedical datasets. It develops an optimized PSO technique to select optimal features and initialize weights for base classifiers in the ensemble model. Experimental results on microarray datasets show the proposed model achieves higher accuracy, true positive rate, and lower error rate compared to traditional feature selection based classification models.
IRJET- Comparative Study of Machine Learning Models for Alzheimer’s Detec...IRJET Journal
This document presents a comparative study of machine learning models for detecting Alzheimer's disease. The study uses MRI data to extract features like brain volume and uses those as inputs to various machine learning models like logistic regression, support vector machines, decision trees, random forests and AdaBoost. The performance of each model is evaluated using metrics like accuracy, sensitivity and specificity. The results show that the random forest model performs the best with the highest prediction rates, indicating it has potential for accurate early detection of Alzheimer's using MRI data.
ICU Mortality Rate Estimation Using Machine Learning and Artificial Neural Ne...IRJET Journal
This study aimed to estimate ICU mortality rates using machine learning models including logistic regression, support vector machines, random forests, decision trees, XGBoost and artificial neural networks (ANN). The researchers applied these models to a dataset of 3999 patients with 42 features related to their medical condition and treatment. The ANN model achieved the highest accuracy of 89% for predicting whether patients would live or die. Confusion matrix, F1-score, accuracy and recall values confirmed the ANN model had strong performance. The researchers concluded ANN was the best model for ICU mortality prediction and could help hospitals optimize resource use and treatment success rates.
IRJET - Alzheimer’s Detection Model Using Machine LearningIRJET Journal
This document discusses a machine learning model for detecting Alzheimer's disease using MRI scans. It proposes using deep learning techniques to accurately detect Alzheimer's and its early stages from structural MRI images. The model would combine MRI data with neuropsychological tests as input for classifying Alzheimer's disease and its prodromal stages. It describes developing a classification model using techniques like data preprocessing, feature extraction, training classifiers like random forests, and selecting the best performing model to use for predictions. The goal is to create a more reliable, cost-effective and fast way to diagnose Alzheimer's compared to conventional expert-based methods.
HEALTH PREDICTION ANALYSIS USING DATA MININGAshish Salve
As we know that health care industry is completely based on assumptions, which after get tested and verified via various tests and patient have to be depend on the doctors knowledge on that topic . so we made a system that uses data mining techniques to predict the health of a person based on various medical test results. so we can predict the health of that person based on that analysis performed by the system.The system currently design only for heart issues, for that we had used Statlog (Heart) Data Set from UCI Machine Learning Repository it includes attributes like age, sex, chest pain type, cholesterol, sugar, outcomes,etc.for training the system. we only need to passed few general inputs in order to generate the prediction and the prediction results from all algorithms are they merged together by calculating there mean value that value shows the actual outcome of the prediction process which entirely works in background
Prediction of Cognitive Imperiment using Deep LearningIRJET Journal
This document proposes using a convolutional neural network (CNN) model to predict cognitive impairment based on MRI data. It describes collecting MRI reports from various sources to create training and test datasets divided into categories for Alzheimer's dementia, healthy controls, and mild cognitive impairment. The CNN model is trained on this data to differentiate between stages of illness. Results showed the CNN approach achieved accuracy of 81.96% for sensitivity, 71.35% for specificity, and 89.72% for precision, outperforming other state-of-the-art methods by around 5%. The proposed system uses CNN to automatically learn features from raw MRI images without need for manual feature extraction, allowing for a more objective and less biased prediction of cognitive impairment.
IRJET - Machine Learning Algorithms for the Detection of DiabetesIRJET Journal
This document discusses machine learning algorithms for detecting diabetes using a diabetes dataset. It evaluates Logistic Regression, Random Forest, Deep Neural Networks (DNN), and a DNN with embeddings for categorical features. The DNN with embeddings achieves near 100% accuracy and an F1 score of 1.0 on the test data, outperforming the other methods. It analyzes the dataset using various algorithms and finds DNN with embeddings to be highly effective at correctly predicting diabetes status in test examples.
An Artificial Neural Network Based Medical Diagnosis of Mental Health DiseasesNathanael Asaam
The document is a research paper that explores using artificial neural networks (ANNs) for medical diagnosis of mental health diseases. Specifically, it focuses on using ANNs to diagnose bipolar disorder and schizophrenia based on patient symptoms. It describes training two ANN algorithms (Perceptron and Adaline) with different activation functions on datasets of patient symptoms and diagnoses. Results showed the Perceptron algorithm was effective at prediction, while Adaline was less accurate.
Health Care Application using Machine Learning and Deep LearningIRJET Journal
This document presents a study on using machine learning and deep learning techniques for healthcare applications like disease prediction. It discusses algorithms like logistic regression, decision trees, random forests, SVMs and deep learning models like VGG16 applied on various disease datasets. For diabetes, heart and liver diseases, ML algorithms were used while CNN models were used for malaria and pneumonia image datasets. Random forest achieved the highest accuracy of 84.81% for diabetes prediction, SVM had 81.57% accuracy for heart disease and random forest was best at 83.33% for liver disease. The VGG16 model attained accuracies of 94.29% and 95.48% for malaria and pneumonia respectively. The study aims to develop an intelligent healthcare application for predicting different
PREDICTION OF MALIGNANCY IN SUSPECTED THYROID TUMOUR PATIENTS BY THREE DIFFER...cscpconf
This document compares three classification methods - artificial neural networks, decision trees, and logistic regression - for predicting malignancy in thyroid tumor patients using a clinical dataset. It describes each method and applies them to a dataset of 259 thyroid tumor patients. The artificial neural network achieved 98% accuracy on the training set and 92% on the validation set. The decision tree method used 150 cases to build a model and achieved 86% accuracy. Logistic regression analysis resulted in 88% accuracy. The artificial neural network was able to accurately predict malignancy and identified important attributes like multiple nodules and family cancer history.
Risk Of Heart Disease Prediction Using Machine LearningIRJET Journal
This document describes a study that uses machine learning algorithms to predict the risk of heart disease. It analyzes a dataset containing characteristics of 270 patients using algorithms like logistic regression, naive Bayes, support vector machine, k-nearest neighbors, decision tree, random forest, XGBoost and artificial neural network. The random forest algorithm achieved the highest prediction accuracy of 95%. The model takes patient attributes as input and outputs a prediction of 0 or 1 indicating the presence or absence of heart disease risk. It aims to help detect risk early to reduce death rates from heart disease, which is a leading cause of death worldwide.
An Introduction To Artificial Intelligence And Its Applications In Biomedical...Jill Brown
1) The document discusses the use of artificial intelligence techniques in biomedical engineering and medicine. It focuses on using AI to analyze medical signals and images to assist clinicians.
2) Key applications discussed include using neural networks and expert knowledge as intelligent agents to aid diagnosis, as well as using AI systems to automatically realign MRI images and identify corresponding slices from different scans.
3) The document also outlines the major components of AI, including problem solving, knowledge representation, and perception, and how AI can be applied through intelligent agents, a task manager, and a communication system to integrate different types of medical data and decision-making approaches.
Classification of physiological diseases using eeg signals and machine learni...eSAT Journals
Abstract
The present day scenario of the modern world is impossible to imagine without automobile. Thus it a primary challenge for automobile industries to design efficient and cost effective engine. The performance of the engine again depends on the type of fuel used, the cooling system, the lubrication system, the range of temperature in which the engine works etc. If the factors are taken care of then the performance of the engine can be improved. In this paper the effect of the fuel assimilating metallic nano-particles is studied. After the combustion of fuel in the combustion chamber the byproducts of combustion (water vapor and carbon dioxide) are further disintegrated, the dissociation of water being an exothermal process adds up to the heat intake of the engine. The food for the engine being heat and not the fuel, it is beneficial in every sense to extract maximum possible amount of heat from the given mass of fuel. This process has both merits and demerits which are shown by the comprehensive study of the fuel used in an internal combustion engine, the chemical process involved in the combustion and the process of exhaust.
Keywords: Nanofluids, Heat Transfer, Conductivity, Knocking and Detonation, Thermal Diffusibility.
Analyzing the behavior of different classification algorithms in diabetes pre...IAESIJAI
Diabetes is one of the deadliest diseases in the world that can lead to stroke, blindness, organ failure, and amputation of lower limbs. Researches state that diabetes can be controlled if it is detected at an early stage. Scientists are becoming more interested in classification algorithms in diagnosing diseases. In this study, we have analyzed the performance of five classification algorithms namely naïve Bayes, support vector machine, multi layer perceptron artificial neural network, decision tree, and random forest using diabetes dataset that contains the information of 2000 female patients. Various metrics were applied in evaluating the performance of the classifiers such as precision, area under the curve (AUC), accuracy, receiver operating characteristic (ROC) curve, f-measure, and recall. Experimental results show that random forest is better than any other classifier in predicting diabetes with a 90.75% accuracy rate.
IRJET- Disease Prediction using Machine LearningIRJET Journal
This document discusses using machine learning techniques to predict diseases based on patient symptoms. Specifically, it proposes using naive bayes, k-nearest neighbors (KNN), and logistic regression algorithms on structured and unstructured hospital data to predict diseases like diabetes, malaria, jaundice, dengue, and tuberculosis. The system is intended to make disease prediction more accessible to end users by analyzing their symptoms without needing to visit a doctor. It aims to improve prediction accuracy by handling both structured and unstructured data using machine learning models.
Pattern recognition system based on support vector machinesAlexander Decker
This document describes a study that uses support vector machines (SVM) to develop quantitative structure-activity relationship (QSAR) models for predicting the anti-HIV activity of 1,3,4-oxadiazole substituted naphthyridine derivatives based on their molecular descriptors. The SVM model achieved a cross-validation R2 value of 0.90 and RMSE of 0.145, outperforming artificial neural network and multiple linear regression models. An external validation on an independent test set found the SVM model had an R value of 0.96 and RMSE of 0.166, demonstrating good predictive ability.
Ataxic person prediction using feature optimized based on machine learning modelIJECEIAES
Ataxic gait monitoring and assessment of neurological disorders belong to important areas that are supported by digital signal processing methods and artificial intelligence (AI) techniques such as machine learning (ML) and deep learning (DL) techniques. This paper uses spatio-temporal data from Kinect sensor to optimize machine learning model to distinguish between ataxic and normal gait. Existing ML-based methodologies fails to establish feature correlation between different gait parameters; thus, exhibit very poor performance. Further, when data is imbalanced in nature the existing ML-based methodologies induces higher false positive. In addressing the research issues this paper introduces an extreme gradient boost (XGBoost)based classifier and enhanced feature optimization (EFO) by modifying the standard cross validation (SCV) mechanism. Experiment outcome shows the proposed ataxic person identification model achieves very good result in comparison with existing ML-based and DL-based ataxic person identification methodologies.
Comparative study of artificial neural network based classification for liver...Alexander Decker
This document presents a comparative study of different artificial neural network (ANN) classification models for predicting liver disease in patients. It evaluates ANN models like backpropagation, radial basis function, self-organizing map, and support vector machine on liver patient data. The support vector machine model achieved the highest accuracy at 99.76% for men data and 97.7% for women data, indicating it may be effective as a predictive tool for liver patients.
A novel salp swarm clustering algorithm for prediction of the heart diseasesnooriasukmaningtyas
Heart diseases have a severe impact on human life and health. Cardiovascular deaths and diseases have increased at a fast rate worldwide. The early prediction of these diseases is necessary to prevent deaths. Now a day; a considerable amount of medical information is available and collected as databases. An efficient technique is required to analyse this data and predict the disease. Clustering can help medical practitioners in diagnosis by classifying the patient’s data collected for a disease. Clustering techniques can analyse such data based on each patient-generated and predict disease. A new prediction model based on salp swarm algorithm and support vector machine is proposed in this research for predicting heart diseases. Salp swarm algorithm is used to select the useful features from the database. Support vector machine classifier is used to predict heart diseases. Results obtained are compared with the other algorithms available in the literature. It is observed that the proposed approach produces better results with accuracy 98.75% and 98.46% with the dataset 1 and 2, respectively. In addition to this, the algorithm converges in significantly less time in comparison to other algorithms. This algorithm might become a perfect supporting tool for medical practitioners.
Similar to Prognosticating Autism Spectrum Disorder Using Artificial Neural Network: Levenberg-Marquardt Algorithm (20)
Feeding plate for a newborn with Cleft Palate.pptxSatvikaPrasad
A feeding plate is a prosthetic device used for newborns with a cleft palate to assist in feeding and improve nutrition intake. From a prosthodontic perspective, this plate acts as a barrier between the oral and nasal cavities, facilitating effective sucking and swallowing by providing a more normal anatomical structure. It helps to prevent milk from entering the nasal passage, thereby reducing the risk of aspiration and enhancing the infant's ability to feed efficiently. The feeding plate also aids in the development of the oral muscles and can contribute to better growth and weight gain. Its custom fabrication and proper fitting by a prosthodontist are crucial for ensuring comfort and functionality, as well as for minimizing potential complications. Early intervention with a feeding plate can significantly improve the quality of life for both the infant and the parents.
This particular slides consist of- what is Pneumothorax,what are it's causes and it's effect on body, risk factors, symptoms,complications, diagnosis and role of physiotherapy in it.
This slide is very helpful for physiotherapy students and also for other medical and healthcare students.
Here is a summary of Pneumothorax:
Pneumothorax, also known as a collapsed lung, is a condition that occurs when air leaks into the space between the lung and chest wall. This air buildup puts pressure on the lung, preventing it from expanding fully when you breathe. A pneumothorax can cause a complete or partial collapse of the lung.
PET CT beginners Guide covers some of the underrepresented topics in PET CTMiadAlsulami
This lecture briefly covers some of the underrepresented topics in Molecular imaging with cases , such as:
- Primary pleural tumors and pleural metastases.
- Distinguishing between MPM and Talc Pleurodesis.
- Urological tumors.
- The role of FDG PET in NET.
INFECTION OF THE BRAIN -ENCEPHALITIS ( PPT)blessyjannu21
Neurological system includes brain and spinal cord. It plays an important role in functioning of our body. Encephalitis is the inflammation of the brain. Causes include viral infections, infections from insect bites or an autoimmune reaction that affects the brain. It can be life-threatening or cause long-term complications. Treatment varies, but most people require hospitalization so they can receive intensive treatment, including life support.
About this webinar: This talk will introduce what cancer rehabilitation is, where it fits into the cancer trajectory, and who can benefit from it. In addition, the current landscape of cancer rehabilitation in Canada will be discussed and the need for advocacy to increase access to this essential component of cancer care.
Let's Talk About It: Breast Cancer (What is Mindset and Does it Really Matter?)bkling
Your mindset is the way you make sense of the world around you. This lens influences the way you think, the way you feel, and how you might behave in certain situations. Let's talk about mindset myths that can get us into trouble and ways to cultivate a mindset to support your cancer survivorship in authentic ways. Let’s Talk About It!
Can Allopathy and Homeopathy Be Used Together in India.pdfDharma Homoeopathy
This article explores the potential for combining allopathy and homeopathy in India, examining the benefits, challenges, and the emerging field of integrative medicine.
LGBTQ+ Adults: Unique Opportunities and Inclusive Approaches to CareVITASAuthor
This webinar helps clinicians understand the unique healthcare needs of the LGBTQ+ community, primarily in relation to end-of-life care. Topics include social and cultural background and challenges, healthcare disparities, advanced care planning, and strategies for reaching the community and improving quality of care.
Dr. David Greene R3 stem cell Breakthroughs: Stem Cell Therapy in CardiologyR3 Stem Cell
Dr. David Greene, founder and CEO of R3 Stem Cell, is at the forefront of groundbreaking research in the field of cardiology, focusing on the transformative potential of stem cell therapy. His latest work emphasizes innovative approaches to treating heart disease, aiming to repair damaged heart tissue and improve heart function through the use of advanced stem cell techniques. This research promises not only to enhance the quality of life for patients with chronic heart conditions but also to pave the way for new, more effective treatments. Dr. Greene's work is notable for its focus on safety, efficacy, and the potential to significantly reduce the need for invasive surgeries and long-term medication, positioning stem cell therapy as a key player in the future of cardiac care.
TEST BANK For Accounting Information Systems, 3rd Edition by Vernon Richardso...rightmanforbloodline
TEST BANK For Accounting Information Systems, 3rd Edition by Vernon Richardson, Verified Chapters 1 - 18, Complete Newest Version
TEST BANK For Accounting Information Systems, 3rd Edition by Vernon Richardson, Verified Chapters 1 - 18, Complete Newest Version
TEST BANK For Accounting Information Systems, 3rd Edition by Vernon Richardson, Verified Chapters 1 - 18, Complete Newest Version
Rate Controlled Drug Delivery Systems, Activation Modulated Drug Delivery Systems, Mechanically activated, pH activated, Enzyme activated, Osmotic activated Drug Delivery Systems, Feedback regulated Drug Delivery Systems systems are discussed here.
Trauma Outpatient Center is a comprehensive facility dedicated to addressing mental health challenges and providing medication-assisted treatment. We offer a diverse range of services aimed at assisting individuals in overcoming addiction, mental health disorders, and related obstacles. Our team consists of seasoned professionals who are both experienced and compassionate, committed to delivering the highest standard of care to our clients. By utilizing evidence-based treatment methods, we strive to help our clients achieve their goals and lead healthier, more fulfilling lives.
Our mission is to provide a safe and supportive environment where our clients can receive the highest quality of care. We are dedicated to assisting our clients in reaching their objectives and improving their overall well-being. We prioritize our clients' needs and individualize treatment plans to ensure they receive tailored care. Our approach is rooted in evidence-based practices proven effective in treating addiction and mental health disorders.
Prognosticating Autism Spectrum Disorder Using Artificial Neural Network: Levenberg-Marquardt Algorithm
1. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 1
Research Article
Prognosticating Autism Spectrum Disorder Using Artificial Neural
Network: Levenberg-Marquardt Algorithm
Avishek Choudhury*, Christopher M Greene
Binghamton University, New York, USA
*Corresponding Author: Avishek Choudhury, Binghamton University, Systems Science and Industrial
Engineering, Engineering Building, L2, Vestal, NY 13902, USA, Tel: +1 (806) 500-8025; E-mail:
achoud13@binghamton.edu
Received: 29 August 2018; Accepted: 05 September 2018; Published: 13 September 2018
Abstract
Autism spectrum condition (ASC) or autism spectrum disorder (ASD) is primarily identified with the help of
behavioral indications encompassing social, sensory and motor characteristics. Although categorized, recurring
motor actions are measured during diagnosis, quantifiable measures that ascertain kinematic physiognomies in the
movement configurations of autistic persons are not adequately studied, hindering the advances in understanding the
etiology of motor mutilation. Subject aspects such as behavioral characters that influences ASD need further
exploration. Presently, limited autism datasets concomitant with screening ASD are available, and a majority of
them are genetic. Hence, in this study, we used a dataset related to autism screening enveloping ten behavioral and
ten personal attributes that have been effective in diagnosing ASD cases from controls in behavior science. ASD
diagnosis is time exhaustive and uneconomical. The burgeoning ASD cases worldwide mandate a need for the fast
and economical screening tool. Our study aimed to implement an artificial neural network with the Levenberg-
Marquardt algorithm to detect ASD and examine its predictive accuracy. Consecutively, develop a clinical decision
support system for early ASD identification.
Keywords: Neural networks; The Levenberg-Marquardt algorithm; Clinal decision support system; Autism
diagnosis
1. Introduction
Autism spectrum disorder (ASD) is a faction of polygenetic evolving brain disarray accompanied with behavioral
and cognitive mutilation [1]. It is a lifelong neurodevelopmental illness depicted by insufficiencies in
communication, interaction and constrained behavior [2]. Though ASD is identified mainly by behavioral and social
2. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 2
physiognomies, autistic individuals often exhibit tainted motor ability such as reduced physical synchronization,
unstable body balance and unusual posture and movement patterns [3-5]. Individuals with ASD show stereotypical
recurring actions, constrained interests, a privation of instinct control, speech insufficiencies, compromised intellect
and social skills compared to typically developing (TD) children [6]. There has been well-established work done in
diagnosing ASD using kinematic physiognomies.
Guha et al. [7] used gesticulation data to measure atypicality variance in facial expressions of children with and
without ASD. Six facial reactions were securitized using information theory, statistical analysis and time- series
modeling. Recently researchers have engrossed in refining and implementing data analytics as a diagnosis tool for
ASD. Being an efficient computational tool machine learning has disclosed its potential for classification in the
various domain [8-9]. Therefore, some literature has employed machine learning methods to identify neural [10]
[11] or behavioral markers [12-13] responsible for discriminating individuals with and without ASD. Stahl et al.
[10] studied the influence of eye gaze to classify infant groups with a high or low risk of getting ASD.
Computational methods such as discriminant functions analysis with an accuracy of 0.61, support vector machine
with an accuracy of 0.64, and linear discrimination analysis with an accuracy of 0.56 were implemented. Bone in
2016 [14] retrieved data from The Autism Diagnostic Interview-Revised and The Social Responsiveness Scale and
applied SVM machine learning classifier to a significant group individual with and without ASD and obtained
promising results and demonstrated that machine learning could be used as a useful tool for ASD diagnosis. Grossi
et al. employed artificial neural networks (ANNs) to develop a predictive model using a dataset comprising of 27
potential pregnancy risk elements in autism development [15]. The artificial neural network produced a predictive
accuracy of 80%. Their study supported and encouraged the use of ANNs as an excellent diagnostic screening tool
for ASD. The highest classification accuracy obtained so far is 96.7% using SVM.
With the aim of developing a clinical decision support system, we implement artificial neural networks with the
Levenberg-Marquardt algorithm on data set that contains ten behavioral and ten personal attributes of adults with
and without ASD.
2. Methodology
This study does not involve any participation of human subjects. We extracted the data from the UCI library. The
data and data description is provided with this paper. It consists of 20 predictors (ten behavioral and ten personal
attributes), one response variable ad 704 instances. The methodology designed for this study can be divided into (a)
data preprocessing, (b) designing the model, and (c) fitting and evaluating the model.
2.1 Data preprocessing
Data preprocessing is one of the most critical steps in all machine learning application. In this study, we did not use
missing data points and partitioned the dataset into training, testing and selection instances. The following pie chart
(figure 1a) details the uses of all the instances in the dataset. The total number of instances is 704, that contains 424
3. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 3
(60.2%) training instances, 140 (19.9%) of selection instances, and 140 (19.9%) of testing instances. The following
pie chart includes all the missing values.
Figure 1: (a) Shows the partitioned data; (b) Shows the number of instances belonging to each class in the dataset.
The pie chart above (figure 1b) displays the partitioning of the dataset (excluding all missing values). The number of
instances with negative Class/ASD (blue) is 222, and the number of instances with positive Class/ASD (purple) is
90. This figure also shows that the data is unbalanced. However, we did not implement any data balancing method.
2.2 Designing the model
In this step we calculated the suitable training algorithm for our dataset and determined the complexity of the model,
that is the optimal number of neurons in the network.
2.2.1 The levenberg–marquardt algorithm: Kenneth Levenberg and Donald Marquardt developed The
Levenberg–Marquardt algorithm (LM) [16-17] that generates a mathematical solution to a problem of minimizing a
non- linear function. We used this algorithm because in the domain of artificial neural-networks it is fast and has
stable convergence. The LM algorithm approaches the second-order training speed without calculating the Hessian
matrix. It holds good when the loss function has the arrangement of a sum of squares. LM is an optimization
algorithm that outperforms simple gradient descent and conjugate gradient methods in a diverse assortment of
problems. LM algorithm follows equation 1 as shown below.
(1)
where J is the Jacobian matrix, T stands for transpose, k is the index of iteration, e is the training error and w is the
weight vector.
The following Table 1 shows a brief description of the parameters used for this algorithm.
a b
4. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 4
Parameters Description Value
Damping parameter factor Damping parameter increase/decrease factor. 10
Minimum parameter increment norm Norm of the parameter increment vector at which training stops. 0.001
Minimum loss increase Minimum loss improvement between two successive iterations 1e-12
Performance goal Goal value for the loss 1e-12
Gradient norm goal Goal value for the norm of the objective function gradient 0.001
Maximum selection loss increase Maximum number of iterations at which the selection loss increases 100
Maximum iterations number Maximum number of iterations to perform the training. 1000
Table 1: The Levenberg-Marquardt algorithm description.
2.2.2 Order selection algorithm: For better performance of the model, we implemented an incremental order
selection algorithm to achieve a model with the best complexity to produce an adequate fit of the data. Incremental
selection order is the naivest order selection algorithm. This method begins with a least order and upsurges the size
of the hidden layer of neurons until a desired order is attained. The order selection algorithm determines the optimal
number of neurons in the network. Incremental order selection enhances the ability of the model to predict the result
with a new data. Two recurrent problems that hinder the design of a neural network are underfitting and overfitting.
The best simplification is attained by designing a model whose complexity is appropriate to produce the best fit
model. Underfitting can occur if the model is too simple, whereas overfitting is defined as the effect of a selection
error burgeoning due to an over-complex model. The next figure 2 illustrates the training and selection errors as a
function of the order of a neural network.
Figure 1: Effect of model complexity on Overfitting and Underfitting.
2.3 Fitting and evaluating the model
After designing the appropriate model, we deployed it on our preprocessed dataset and computed the following
performance measures: (a) Accuracy, (b) Sum squared error, (c) mean squared error, (d) root mean squared error, (e)
5. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 5
normalized squared error, (f) cross-entropy error, (g) Minkowski error, and (h) weighted squared error. Additionally,
the area under the ROC curve, miss-classification, cumulative gain, lift chart and model loss index were observed
for better understanding and evaluation of the model’s performance.
3. Results
Implementing the Levenberg–Marquardt algorithm and incremental order selection enabled the proposed model to
produce a classification accuracy of 98.38%. The following table 2 shows the training results obtained by the
Levenberg-Marquardt algorithm. They include final states from the neural network, the loss function, and the
training algorithm. The number of iterations needed to converge is found to be zero, which infers that the training
algorithm did not modify the state of the neural network.
Parameters Value
Final parameters norm 14.5
Final loss 9.89e-05
Final selection loss 0.00122
Final gradient norm 0.00036
Iteration number 0
Elapsed time 0
Stopping criterion Gradient norm goal
Table 2: The result obtained from the Levenberg-Marquardt algorithm.
The optimal number of neurons was calculated to be 1. The figure 3 displays the loss history for the different subsets
during the incremental order selection process. The blue line represents the training loss, and the red line symbolizes
the selection loss.
Figure 3: Loss history for different subsets obtained during incremental order selection algorithm.
6. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 6
Parameters Value
Optimal order 1
Optimum training loss 8.2344e-05
Optimum selection loss 0.00313
Iteration number 20
Table 3: Final output obtained from incremental order selection algorithm.
3.1 Testing errors
This task measures all the losses of the model. It considers every used instance and evaluates the model for each use.
The table 4 below shows all the training, testing and selection errors of the data.
Errors Training Selection Testing
Sum squared error 0.0032 0.0796 0.3243
Mean squared error 1.7360e- 05
0.0012 0.0052
Root mean squared error 0.0041 0.0358 0.0723
Normalized squared error 7.0236e-05
0.0052 0.0210
Cross- entropy error 0.0036 0.0134 0.0188
Minkowski error 0.0459 0.2236 0.4722
Weighted squared error 9.8919e- 05
0.0122 0.0241
Table 4: All performance measures (Testing errors).
3.2 Confusion table
This section shows the correct and miss classifications made by the model on the testing dataset. The table 5 below
is termed as a confusion matrix. The rows in the table denotes the target variables, whereas the columns represent
the output classes belonging to the testing dataset. The diagonal cells depict the correctly classified, and the off-
diagonal cells indicates the misclassified instances.
The decision threshold is considered to be 0.5. The table shows 61 correctly classified instances and one
misclassified instance.
Instances Predicted positive Predicted negative
Actual positive 33 0
Actual negative 1 28
Table 5: Confusion table showing positive and negative classification.
7. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 7
Classification accuracy was then calculated from the confusion matrix using the following equation 2:
[(Correct prediction) / (Total predicted)] * 100 (2)
The following part of this section will evaluate the model’s performance based on ROC, Cumulative gain, Lift plot
and Loss index.
3.3 Receiver Operating Characteristic
A good way to analyze the loss is by plotting a ROC (Receiver Operating Characteristic) curve which is a graphical
illustration of how good the classifier distinguishes between the two dissimilar classes. This capacity for
discrimination is measured by calculating the area under the curve. ROC was found to be 1 with an optimal
threshold of 0.882. The figure 4 below shows the ROC curve obtained and the blue shaded region is the AUC.
Figure 4: ROC plot.
3.4 Cumulative Gain
The cumulative gain analysis is a pictorial representation that illustrates the benefit of implementing a predictive
model as opposed to randomness. The baseline represents the results that would be obtained without using a model.
The positive cumulative gain which shows on the y-axis the percentage of positive instances found against the
percentage of the population, which is represented on the x-axis.
Similarly, the negative cumulative gain represents the fraction of the negative instances found against the
population. The following figure 5 shows the results of the analysis in this case. The blue line represents the positive
cumulative gain, the red line shows the negative cumulative gain, and the grey line signifies the random cumulative
gain. Greater separation between the positive and negatives cumulative gain charts, indicates better performance of
the classifier.
8. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 8
Figure 5: Cumulative plot.
The maximum gain score of a cumulative gain is obtained by the computing the maximum difference between the
positive and the negative cumulative gain, i.e., it is the point where the percentage of positive instances found is
maximized, and the percentage of negative instances found is minimized. If we have a perfect model, this score
takes the value 1. The next table 6 shows the score in this case and the instances ratio for which it is reached.
Parameters Value
Instance ratio 0.55
Maximum gain score 0.97
Table 6: Maximum gain score.
3.5 Lift plot
A lift plot provides a visual aid for evaluating a predictive model loss. It consists of a lift curve and a baseline. Lift
curve represents the ratio between the positive events using a model and without using it. Baseline represents
randomness, i.e., not using a model. The figure 6 below shows the lift plot obtained for this study. The x-axis
displays the percentage of instances studied while the y-axis displays the ratio between the results predicted by the
model and the results without the model.
Figure 6: Lift plot.
9. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 9
3.6 Loss index
The loss index plays an essential role in the use of a neural network. It delineates the errand that the neural network
is designed to perform and provides a gauge of the quality of the desired learning. The loss index varies with
different applications. We implemented the weighted squared error as the error method which is especially useful
when the data set has unbalanced targets. That is, there are too few positives when compared to the negatives, or
vice versa. The following table 7 shows the exponent in error between the outputs from the neural network and the
targets in the dataset.
Variable Value
Positive weight 2.72
Negative weight 1
Table 7: Exponent in error between the output from the ANN and the target variable.
4. Conclusion
To our knowledge, this is the first attempt to examine identification of autism using the Levenberg-Marquardt
algorithm and incremental order selection in the field of artificial neural networks. Moreover, our model produced
the highest classification accuracy of 98.38%. Through this study, we highlighted the importance of the training
algorithm, order selection algorithm and selecting the required loss index to deal with unbalanced data. Based on the
observations and evaluation of the proposed model, it can be inferred that Neural network with the Levenberg-
Marquardt algorithm and incremental order selection is an appropriate tool for diagnosing ASD and can be deployed
as a clinical decision support system.
References
1. Rapin, Tuchman R. Autism: definition, neurobiology, screening, diagnosis. Pediatr Clin North Am. Pediatr Clin
North Am (2008): 1129-1146.
2. Baihua Li JMEG. American Psychiatric Association, Diagnostic and statistical manual of mental disorders.
PLOS One, American Psychiatric Publishing 12 (2017): 1-19.
3. Gowen E, Hamilton A. Motor abilities in autism: a review using a computational context. Journal of autism and
developmental disorders 43 (2013): 323-344.
4. Fournier K, Hass C, Naik S, et al. Motor coordination in autism spectrum disor- ders: A synthesis and meta-
analysis. Journal of autism and developmental disorders 40 (2010): 1227-1240.
5. Bhat, Landa R, Galloway J. Current perspectives on motor functioning in infants, children, and adults with
autism spectrum disorders. Physical Therapy 91 (2011): 1116-1129.
6. Kang S, Reilly MO, Rojeski L, et al. Effects of tangible and social reinforcers on skill acquisition, stereotyped
behavior, and task engagement in three children with autism spectrum disorders. Res Dev Disability 34 (2013):
739-744.
10. J Bioinform Syst Biol 2018; 1 (1): 001-011 DOI: 10.26502/fjbsb001
Journal of Bioinformatics and Systems Biology - Vol. 1 No. 1 - Sep 2018. 10
7. Guha T, Yang Z, Ramakrishna A, et al. On quantifying facial expres- sion-related atypicality of children with
autism spectrum disorder. IEEE Int Conf Acoustics, Seech and Signal Processing (ICASSP) 803-807.
8. Mitchell T. Machine learning. Boston: McGraw-Hill 1997.
9. Kotsiantis S, Zaharakis I, Pintelas P. Supervised machine learning: A review of classification techniques. 2007.
10. Stahl D, Pickles A, Elsabbagh M, et al. Novel machine learning methods for ERP analysis: A validation from
research on infants at risk for autism. Developmental neuropsychology 37 (2012): 274-298.
11. Chanel G, Pichon S, Conty L, et al. Classification of autistic individuals and controls using cross-task
characterization of fMRI activity. NeuroImage: Clinical 10 (2016): 78-88.
12. Kosmicki J, Sochat V, Duda M et al. Searching for a minimal set of behaviors for autism detection through
feature selection-based machine learning. Translational psychiatry 5 (2015): 514.
13. Wall D, Kosmicki J, Deluca T, et al. Use of machine learning to shorten observation- based screening and
diagnosis of autism. Translational psychiatry 2 (2012): 100.
14. Bone D, Bishop S, Black M, et al. Use of machine learning to improve autism screening and diagnostic
instruments: effectiveness, efficiency, and multi-instrument fusion. Journal of Child Psychology and Psychiatry
57 (2016): 927-937.
15. Grossi E, Veggo F, Narzisi A, et al. Pregnancy risk factors in autism: a pilot study with artificial neural
networks. Pediatric research 79 (2016): 339-347.
16. Levenberg K. A method for the solution of certain problems in least squares. Quarterly of Applied
Mathematics. 5 (1944): 164-168.
17. Marquardt D. An algorithm for least-squares estimation of nonlinear parameters. SIAM Journal on Applied
Mathematics. 11 (1963): 431-441.
This article is an open access article distributed under the terms and conditions of the
Creative Commons Attribution (CC-BY) license 4.0
Citation: Avishek Choudhury, Christopher M Greene. Prognosticating Autism Spectrum Disorder Using Artificial
Neural Network: Levenberg-Marquardt Algorithm. Journal of Bioinformatics and Systems Biology 1 (2018): 001-010.