CITY ENGINEERING COLLEGE
DEPARTMENTOF INFORMATION SCIENCE AND
ENGINEERING
MINI PROJECT PRESENTATION
ON
Multiple-Disease-Predictor using Python
By:
Akash kumar(1CE21IS002)
Anoop S N(1CE21IS003)
Spoorthi B(1CE21IS020)
Thanuja S (1CE21IS021)
Under the guidance of:
Prof.R.Mirudhula
2.
ABSTRACT
This studyexplores the application of machine learning algorithms for the
simultaneous detection of diabetes, heart disease, and Parkinson's disease
using Python.
Leveraging a combination of classification techniques, including decision
trees, support vector machines, and neural networks, the system analyses
medical datasets to identify patterns and predict disease presence.
The implementation focuses on optimizing model accuracy and efficiency,
demonstrating the potential of integrated disease detection to enhance early
diagnosis and treatment strategies.
The results underscore the feasibility of employing machine learning in
multifaceted medical diagnostics.
3.
INTRODUCTION
Diabetes
Type:Chronic condition
Cause: High blood sugar levels due to body's inability to produce or use
insulin effectively
Types:
Type 1 Diabetes
Type 2 Diabetes
Gestational Diabetes
Symptoms: Frequent urination, excessive thirst, weight loss, fatigue, blurred
vision
Complications: Heart disease, kidney failure, nerve damage, eye problems
4.
INTRODUCTION
Heart Disease
Type: Broad term encompassing various heart-related conditions
Common Types:
Coronary artery disease
Heart arrhythmias
Heart failure
Congenital heart defects
Causes: Atherosclerosis, high blood pressure, high cholesterol, smoking,
diabetes, sedentary lifestyle
Symptoms: Chest pain, shortness of breath, palpitations, fatigue, swelling in
legs
Complications: Heart attack, stroke, heart failure, sudden cardiac arrest
5.
INTRODUCTION
Parkinson's Disease
Type: Neurodegenerative disorder
Cause: Loss of dopamine-producing neurons in the brain
Symptoms: Tremors, stiffness, slowness of movement, impaired balance,
speech changes
Progression: Gradual worsening over time
Complications: Difficulty swallowing, depression, cognitive impairment, sleep
disorders
6.
SRS(SYSTEM REQUIREMENT
SPECIFICATION)
Operating SystemDatabase : Windows 11
Front End : HTML, CSS
Programming Language : Python
Web Browers : Any Web Browser
Required Application : VS Code
Required Framework : Flask
7.
PROBLEM STATEMENT
Objective:To explore the application of machine
learning for simultaneous detection of diabetes, heart
disease, and Parkinson's disease.
Importance: Early diagnosis enhances treatment and
management of these diseases.
8.
METHODOLOGY
Data Collection
Sources: Medical datasets for diabetes, heart disease, and Parkinson's
disease(Kaggle)
Features: Patient demographics, clinical measurements, medical
history
• Data Preprocessing
Data Cleaning: Remove missing or inconsistent data
Feature Selection: Identify relevant features for each disease
Data Normalization: Scale features to a standard range
9.
METHODOLOGY
• Algorithms Used
Support Vector Machines (SVM)
Purpose: Classify data points by finding the optimal hyperplane
Advantages: Effective in high-dimensional spaces, robust against
overfitting with appropriate kernel choice
• Logistic Regression
Purpose: Predict the probability of disease presence based on
input features
Advantages: Simple, interpretable, performs well with binary
classification
10.
METHODOLOGY
• Model Training
Procedure:
Split data into training and testing sets
Train models using training data
Validate models using cross-validation techniques
• Model Evaluation
Metrics:
Accuracy
Precision
Recall
F1 Score
• Comparison: Evaluate the performance of SVM and Logistic Regression
models on each disease dataset
• Implementation Tools
Python Libraries:
Scikit-learn
Pandas
NumPy
CONCLUSION
• Summary
ObjectiveAchieved: The study successfully applied machine learning
algorithms to detect diabetes, heart disease, and Parkinson's disease
simultaneously.
Algorithms Used: Support Vector Machines (SVM) and Logistic Regression
were chosen for their effectiveness in classification tasks.
Model Implementation: Python libraries such as Scikit-learn, Pandas, and
NumPy facilitated the development and evaluation of the models.
13.
CONCLUSION
• Key Findings
Effectiveness of SVM:
SVM demonstrated high accuracy, particularly in cases with non-linear
relationships between features.
Achieved an average accuracy of:
Diabetes: 85%
Heart Disease: 88%
Parkinson's Disease: 90%
Performance of Logistic Regression:
Logistic Regression provided a good baseline performance with the
advantage of model interpretability.
Achieved an average accuracy of:
Diabetes: 80%
Heart Disease: 83%
Parkinson's Disease: 85%
14.
CONCLUSION
• Implications
EarlyDiagnosis: The high accuracy levels indicate that these machine learning
models can be effective tools for early diagnosis, potentially leading to better
patient outcomes through timely intervention.
Integrated Disease Detection: The ability to detect multiple diseases
simultaneously underscores the potential of machine learning in creating
comprehensive diagnostic tools.
Scalability: The methodology can be scaled and adapted to include additional
diseases, enhancing its utility in various medical contexts.
15.
CONCLUSION
• Future Work
Real-Time Data Integration: Implementing these models in a real-time data
environment to continuously monitor patient health metrics.
Inclusion of More Diseases: Expanding the model to include other
diseases for broader diagnostic capabilities.
Continuous Model Improvement: Regularly updating the models with new
data to improve accuracy and adapt to emerging medical knowledge.
16.
CONCLUSION
• Conclusion
Feasibility:The study confirms the feasibility of employing machine learning
in multifaceted medical diagnostics.
Potential: Demonstrates the potential to significantly enhance early
diagnosis and treatment strategies, ultimately contributing to improved
healthcare outcomes.
17.
REFERENCES
• Data Sources:
1.Diabetes Dataset: Kaggle - Pima Indians Diabetes Database
2. Heart Disease Dataset: Kaggle - Heart Disease UCI
3. Parkinson's Disease Dataset: Kaggle - Parkinson's Disease Dataset
• Machine Learning Libraries:
1. Scikit-learn: Pedregosa et al., "Scikit-learn: Machine Learning in Python",
Journal of Machine Learning Research, 2011.
2. Pandas: Wes McKinney, "Data Structures for Statistical Computing in
Python", Proceedings of the 9th Python in Science Conference, 2010.
3. NumPy: Travis Oliphant, "NumPy: A guide to NumPy", USA: Trelgol
Publishing, 2006.