SlideShare a Scribd company logo
1 of 1
Development of a Scalable Neural Network Platform for Predictive Metabonomics
Marilyn B. Arceo 1,3 and Grady Hanrahan, Ph.D2,3
1Department of Computer Science, California Lutheran University, Thousand Oaks, CA 91360
2Department of Chemistry, California Lutheran University, Thousand Oaks, CA 91360
3Hugh and Hazel Darling Center for Applied Scientific Computing, California Lutheran University, Thousand Oaks, CA 91360
Artificial neural networks (ANNs) are robust computational methodologies
consisting of interconnected processing elements called nodes or neurons that
work together to produce an output function. Inspired by the human central
nervous system, neural networks have been used in various disciplines and fields.
ANNs have been effective computational method that has been used in chemical
applications and in medicine. The goal is to deviate from the black box method of
neural networks that most commercial software uses and create a white box
method that allows the user to modify the mathematical components of the
algorithm. This allows the user to fully have control of the architecture of the
neural network for the given problem at hand. In addition, we would use particle
swarm optimization (PSO) as a training technique for ANN. PSO has become
more effective and accurate in decision making. The members of the swarms
adapt their behavior due to other swarm member behaviors and environment;
thus, creating self-evolving and self-organizing behavior. Using this, we will be
able to make better predictions for each given data set.
INTRODUCTION
Black Box Model vs. White Box Model
The black box model for ANNs have been concealed and their inner workings have
been not available to the public due to commercial software. As a result, white box
models have been developed in order to have the freedom to create a system where
the components and logic of the neural network can be available. This gives the
neural network the ability to solve nonlinear problems instead of being limited to
linear situations.
PSO
The PSO algorithm that will be used for training the ANNs is defined by the
direction and movement of each particle through the search space, by updating its:
velocity: (1)
and position: (2)
METHODS AND THEORY
The implementation of the neural network was through OpenNN C++ library in
Microsoft Visual Studio 2010. The neural network is fully capable of being modified
by the user in order to adjust the neural network architecture for each different set of
metabonomic data. As a result, we were able to run different data sets and develop
overall predictions.
RESULTS & DISCUSSION
RESULTS & DISCUSSION
We would like to thank the Hugh and Hazel Darling Foundation for funding this
project. We would also like to acknowledge the Office of Undergraduate
Research and Creative Scholarship at California Lutheran University. Finally, we
thank Dr. Craig Reinhart for his valuable input.
ACKNOWLEDGEMENTS
ALGORITHMIC FRAMEWORK
1. Dayhoff, J.E., DeLeo, J.M., Artificial Neural Networks Opening the Black
Box, Cancer, 2001, 91, 1615-1635.
2. Hanrahan, G Computational Neural Networks Driving Complex Analytical
Problem Solving, Anal. Chem, 2010, 82, 4307-4313.
3. Lopez, R., OpenNN: Open Neural Networks Library (Version 0.9) [software].,
2012, Retrieved from http://flood.sourceforge.net.
REFERENCES
Fig. 3. The architecture of a white box model for ANNs where PSO will be used to
train the data.
Fig. 4. Pseudo code for ANN framework and logic for the input layer,
hidden layer and output layer of the algorithm.
In the future, we would like to use ANNs for structural characterization and trend
analysis, structure activity relationships, and for instrument optimization. A
graphical user interface for the neural network would help students or instructors
to easily use and create a neural network based off in their needs.
FUTURE WORK
Fig. 1. Vector representation of PSO velocity and position updates in a two-
dimensional space. At each iteration k a particle updates its velocity and
position using Equations (1) and (2), respectively. This process allows all
particles in the swarm to update their and . The solution vector (the position
vector of the swarm of the) gives the optimal set of values for the PSO
parameters s1, s2, and wk .
Fig. 2. Visual representation of the architecture of a black box model for
ANN. The input layer (representing the data), the hidden layer (where the
transformation of our inputs through a function occurs), and the output
layer (representing the predictions computed from hidden layer).
While the network = not trained
for i in each data[i] in dataset
y = Σ wi * datai - T
a = 1 / ( 1 + e –datai) #activation
function
if y != target #doesn’t reach
target value
network = not trained
for each wi
modify weight_connection(wi)
if y = target and reaches end of
data set
network = trained
if y = target #if it reaches the
target value
repeat for data[i+1]
)(111
i
k
i
k
i
kk
i
k vprsvwv 
i
k
i
k
i
k vxx 11  
Fig. 5. Sample prompt from a run for the ANN algorithm that was written in C++
and implemented in Microsoft Visual Studio.

More Related Content

What's hot

Targeted Visual Content Recognition Using Multi-Layer Perceptron Neural Network
Targeted Visual Content Recognition Using Multi-Layer Perceptron Neural NetworkTargeted Visual Content Recognition Using Multi-Layer Perceptron Neural Network
Targeted Visual Content Recognition Using Multi-Layer Perceptron Neural Networkijceronline
 
Handwritten digits recognition report
Handwritten digits recognition reportHandwritten digits recognition report
Handwritten digits recognition reportSwayamdipta Saha
 
Handwritten Digit Recognition
Handwritten Digit RecognitionHandwritten Digit Recognition
Handwritten Digit Recognitionijtsrd
 
Nature Inspired Reasoning Applied in Semantic Web
Nature Inspired Reasoning Applied in Semantic WebNature Inspired Reasoning Applied in Semantic Web
Nature Inspired Reasoning Applied in Semantic Webguestecf0af
 
International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)ijceronline
 
Unsupervised learning
Unsupervised learningUnsupervised learning
Unsupervised learningamalalhait
 
Emotion detection using cnn.pptx
Emotion detection using cnn.pptxEmotion detection using cnn.pptx
Emotion detection using cnn.pptxRADO7900
 
Pattern Recognition using Artificial Neural Network
Pattern Recognition using Artificial Neural NetworkPattern Recognition using Artificial Neural Network
Pattern Recognition using Artificial Neural NetworkEditor IJCATR
 
Gradient-Based Meta-Learning with Learned Layerwise Metric and Subspace
Gradient-Based Meta-Learning with Learned Layerwise Metric and SubspaceGradient-Based Meta-Learning with Learned Layerwise Metric and Subspace
Gradient-Based Meta-Learning with Learned Layerwise Metric and SubspaceYoonho Lee
 
A Time Series ANN Approach for Weather Forecasting
A Time Series ANN Approach for Weather ForecastingA Time Series ANN Approach for Weather Forecasting
A Time Series ANN Approach for Weather Forecastingijctcm
 
Artificial Neural Network and its Applications
Artificial Neural Network and its ApplicationsArtificial Neural Network and its Applications
Artificial Neural Network and its Applicationsshritosh kumar
 
Model of Differential Equation for Genetic Algorithm with Neural Network (GAN...
Model of Differential Equation for Genetic Algorithm with Neural Network (GAN...Model of Differential Equation for Genetic Algorithm with Neural Network (GAN...
Model of Differential Equation for Genetic Algorithm with Neural Network (GAN...Sarvesh Kumar
 
Cat and dog classification
Cat and dog classificationCat and dog classification
Cat and dog classificationomaraldabash
 
GUI based handwritten digit recognition using CNN
GUI based handwritten digit recognition using CNNGUI based handwritten digit recognition using CNN
GUI based handwritten digit recognition using CNNAbhishek Tiwari
 
Neural Network Applications In Machining: A Review
Neural Network Applications In Machining: A ReviewNeural Network Applications In Machining: A Review
Neural Network Applications In Machining: A ReviewAshish Khetan
 
Comparison of Learning Algorithms for Handwritten Digit Recognition
Comparison of Learning Algorithms for Handwritten Digit RecognitionComparison of Learning Algorithms for Handwritten Digit Recognition
Comparison of Learning Algorithms for Handwritten Digit RecognitionSafaa Alnabulsi
 

What's hot (20)

Targeted Visual Content Recognition Using Multi-Layer Perceptron Neural Network
Targeted Visual Content Recognition Using Multi-Layer Perceptron Neural NetworkTargeted Visual Content Recognition Using Multi-Layer Perceptron Neural Network
Targeted Visual Content Recognition Using Multi-Layer Perceptron Neural Network
 
Handwritten digits recognition report
Handwritten digits recognition reportHandwritten digits recognition report
Handwritten digits recognition report
 
Handwritten Digit Recognition
Handwritten Digit RecognitionHandwritten Digit Recognition
Handwritten Digit Recognition
 
Nature Inspired Reasoning Applied in Semantic Web
Nature Inspired Reasoning Applied in Semantic WebNature Inspired Reasoning Applied in Semantic Web
Nature Inspired Reasoning Applied in Semantic Web
 
International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)International Journal of Computational Engineering Research(IJCER)
International Journal of Computational Engineering Research(IJCER)
 
Unsupervised learning
Unsupervised learningUnsupervised learning
Unsupervised learning
 
Emotion detection using cnn.pptx
Emotion detection using cnn.pptxEmotion detection using cnn.pptx
Emotion detection using cnn.pptx
 
Project Report -Vaibhav
Project Report -VaibhavProject Report -Vaibhav
Project Report -Vaibhav
 
Pattern Recognition using Artificial Neural Network
Pattern Recognition using Artificial Neural NetworkPattern Recognition using Artificial Neural Network
Pattern Recognition using Artificial Neural Network
 
Gradient-Based Meta-Learning with Learned Layerwise Metric and Subspace
Gradient-Based Meta-Learning with Learned Layerwise Metric and SubspaceGradient-Based Meta-Learning with Learned Layerwise Metric and Subspace
Gradient-Based Meta-Learning with Learned Layerwise Metric and Subspace
 
A Time Series ANN Approach for Weather Forecasting
A Time Series ANN Approach for Weather ForecastingA Time Series ANN Approach for Weather Forecasting
A Time Series ANN Approach for Weather Forecasting
 
Neural network
Neural networkNeural network
Neural network
 
Artificial Neural Network and its Applications
Artificial Neural Network and its ApplicationsArtificial Neural Network and its Applications
Artificial Neural Network and its Applications
 
Model of Differential Equation for Genetic Algorithm with Neural Network (GAN...
Model of Differential Equation for Genetic Algorithm with Neural Network (GAN...Model of Differential Equation for Genetic Algorithm with Neural Network (GAN...
Model of Differential Equation for Genetic Algorithm with Neural Network (GAN...
 
Cat and dog classification
Cat and dog classificationCat and dog classification
Cat and dog classification
 
Neural network
Neural networkNeural network
Neural network
 
GUI based handwritten digit recognition using CNN
GUI based handwritten digit recognition using CNNGUI based handwritten digit recognition using CNN
GUI based handwritten digit recognition using CNN
 
Neural Network Applications In Machining: A Review
Neural Network Applications In Machining: A ReviewNeural Network Applications In Machining: A Review
Neural Network Applications In Machining: A Review
 
Comparison of Learning Algorithms for Handwritten Digit Recognition
Comparison of Learning Algorithms for Handwritten Digit RecognitionComparison of Learning Algorithms for Handwritten Digit Recognition
Comparison of Learning Algorithms for Handwritten Digit Recognition
 
Neural networks
Neural networksNeural networks
Neural networks
 

Similar to Predictive Metabonomics

Artificial Neural Networks.pdf
Artificial Neural Networks.pdfArtificial Neural Networks.pdf
Artificial Neural Networks.pdfBria Davis
 
20141003.journal club
20141003.journal club20141003.journal club
20141003.journal clubHayaru SHOUNO
 
Dissertation character recognition - Report
Dissertation character recognition - ReportDissertation character recognition - Report
Dissertation character recognition - Reportsachinkumar Bharadva
 
Fuzzy Logic Final Report
Fuzzy Logic Final ReportFuzzy Logic Final Report
Fuzzy Logic Final ReportShikhar Agarwal
 
modeling-a-perceptron-neuron-using-verilog-developed-floating-point-numbering...
modeling-a-perceptron-neuron-using-verilog-developed-floating-point-numbering...modeling-a-perceptron-neuron-using-verilog-developed-floating-point-numbering...
modeling-a-perceptron-neuron-using-verilog-developed-floating-point-numbering...RioCarthiis
 
NEURAL NETWORK FOR THE RELIABILITY ANALYSIS OF A SERIES - PARALLEL SYSTEM SUB...
NEURAL NETWORK FOR THE RELIABILITY ANALYSIS OF A SERIES - PARALLEL SYSTEM SUB...NEURAL NETWORK FOR THE RELIABILITY ANALYSIS OF A SERIES - PARALLEL SYSTEM SUB...
NEURAL NETWORK FOR THE RELIABILITY ANALYSIS OF A SERIES - PARALLEL SYSTEM SUB...IAEME Publication
 
Student intervention detection using deep learning technique
Student intervention detection using deep learning techniqueStudent intervention detection using deep learning technique
Student intervention detection using deep learning techniqueVenkat Projects
 
IRJET- The Essentials of Neural Networks and their Applications
IRJET- The Essentials of Neural Networks and their ApplicationsIRJET- The Essentials of Neural Networks and their Applications
IRJET- The Essentials of Neural Networks and their ApplicationsIRJET Journal
 
NeuralProcessingofGeneralPurposeApproximatePrograms
NeuralProcessingofGeneralPurposeApproximateProgramsNeuralProcessingofGeneralPurposeApproximatePrograms
NeuralProcessingofGeneralPurposeApproximateProgramsMohid Nabil
 
Towards neuralprocessingofgeneralpurposeapproximateprograms
Towards neuralprocessingofgeneralpurposeapproximateprogramsTowards neuralprocessingofgeneralpurposeapproximateprograms
Towards neuralprocessingofgeneralpurposeapproximateprogramsParidha Saxena
 
Artificial Neural Networks: Applications In Management
Artificial Neural Networks: Applications In ManagementArtificial Neural Networks: Applications In Management
Artificial Neural Networks: Applications In ManagementIOSR Journals
 
Final cnn shruthi gali
Final cnn shruthi galiFinal cnn shruthi gali
Final cnn shruthi galiSam Ram
 
Student intervention detection using deep learning technique
Student intervention detection using deep learning techniqueStudent intervention detection using deep learning technique
Student intervention detection using deep learning techniqueVenkat Projects
 

Similar to Predictive Metabonomics (20)

Artificial Neural Networks.pdf
Artificial Neural Networks.pdfArtificial Neural Networks.pdf
Artificial Neural Networks.pdf
 
20141003.journal club
20141003.journal club20141003.journal club
20141003.journal club
 
Dissertation character recognition - Report
Dissertation character recognition - ReportDissertation character recognition - Report
Dissertation character recognition - Report
 
Fuzzy Logic Final Report
Fuzzy Logic Final ReportFuzzy Logic Final Report
Fuzzy Logic Final Report
 
modeling-a-perceptron-neuron-using-verilog-developed-floating-point-numbering...
modeling-a-perceptron-neuron-using-verilog-developed-floating-point-numbering...modeling-a-perceptron-neuron-using-verilog-developed-floating-point-numbering...
modeling-a-perceptron-neuron-using-verilog-developed-floating-point-numbering...
 
Artificial neural network
Artificial neural networkArtificial neural network
Artificial neural network
 
OSPEN: an open source platform for emulating neuromorphic hardware
OSPEN: an open source platform for emulating neuromorphic hardwareOSPEN: an open source platform for emulating neuromorphic hardware
OSPEN: an open source platform for emulating neuromorphic hardware
 
Artificial Neural Networks
Artificial Neural NetworksArtificial Neural Networks
Artificial Neural Networks
 
P2-Artificial.pdf
P2-Artificial.pdfP2-Artificial.pdf
P2-Artificial.pdf
 
NEURAL NETWORK FOR THE RELIABILITY ANALYSIS OF A SERIES - PARALLEL SYSTEM SUB...
NEURAL NETWORK FOR THE RELIABILITY ANALYSIS OF A SERIES - PARALLEL SYSTEM SUB...NEURAL NETWORK FOR THE RELIABILITY ANALYSIS OF A SERIES - PARALLEL SYSTEM SUB...
NEURAL NETWORK FOR THE RELIABILITY ANALYSIS OF A SERIES - PARALLEL SYSTEM SUB...
 
Student intervention detection using deep learning technique
Student intervention detection using deep learning techniqueStudent intervention detection using deep learning technique
Student intervention detection using deep learning technique
 
Neural Networks
Neural NetworksNeural Networks
Neural Networks
 
IRJET- The Essentials of Neural Networks and their Applications
IRJET- The Essentials of Neural Networks and their ApplicationsIRJET- The Essentials of Neural Networks and their Applications
IRJET- The Essentials of Neural Networks and their Applications
 
NeuralProcessingofGeneralPurposeApproximatePrograms
NeuralProcessingofGeneralPurposeApproximateProgramsNeuralProcessingofGeneralPurposeApproximatePrograms
NeuralProcessingofGeneralPurposeApproximatePrograms
 
Towards neuralprocessingofgeneralpurposeapproximateprograms
Towards neuralprocessingofgeneralpurposeapproximateprogramsTowards neuralprocessingofgeneralpurposeapproximateprograms
Towards neuralprocessingofgeneralpurposeapproximateprograms
 
DATA SCIENCE
DATA SCIENCEDATA SCIENCE
DATA SCIENCE
 
Artificial Neural Networks: Applications In Management
Artificial Neural Networks: Applications In ManagementArtificial Neural Networks: Applications In Management
Artificial Neural Networks: Applications In Management
 
Bc34333339
Bc34333339Bc34333339
Bc34333339
 
Final cnn shruthi gali
Final cnn shruthi galiFinal cnn shruthi gali
Final cnn shruthi gali
 
Student intervention detection using deep learning technique
Student intervention detection using deep learning techniqueStudent intervention detection using deep learning technique
Student intervention detection using deep learning technique
 

Predictive Metabonomics

  • 1. Development of a Scalable Neural Network Platform for Predictive Metabonomics Marilyn B. Arceo 1,3 and Grady Hanrahan, Ph.D2,3 1Department of Computer Science, California Lutheran University, Thousand Oaks, CA 91360 2Department of Chemistry, California Lutheran University, Thousand Oaks, CA 91360 3Hugh and Hazel Darling Center for Applied Scientific Computing, California Lutheran University, Thousand Oaks, CA 91360 Artificial neural networks (ANNs) are robust computational methodologies consisting of interconnected processing elements called nodes or neurons that work together to produce an output function. Inspired by the human central nervous system, neural networks have been used in various disciplines and fields. ANNs have been effective computational method that has been used in chemical applications and in medicine. The goal is to deviate from the black box method of neural networks that most commercial software uses and create a white box method that allows the user to modify the mathematical components of the algorithm. This allows the user to fully have control of the architecture of the neural network for the given problem at hand. In addition, we would use particle swarm optimization (PSO) as a training technique for ANN. PSO has become more effective and accurate in decision making. The members of the swarms adapt their behavior due to other swarm member behaviors and environment; thus, creating self-evolving and self-organizing behavior. Using this, we will be able to make better predictions for each given data set. INTRODUCTION Black Box Model vs. White Box Model The black box model for ANNs have been concealed and their inner workings have been not available to the public due to commercial software. As a result, white box models have been developed in order to have the freedom to create a system where the components and logic of the neural network can be available. This gives the neural network the ability to solve nonlinear problems instead of being limited to linear situations. PSO The PSO algorithm that will be used for training the ANNs is defined by the direction and movement of each particle through the search space, by updating its: velocity: (1) and position: (2) METHODS AND THEORY The implementation of the neural network was through OpenNN C++ library in Microsoft Visual Studio 2010. The neural network is fully capable of being modified by the user in order to adjust the neural network architecture for each different set of metabonomic data. As a result, we were able to run different data sets and develop overall predictions. RESULTS & DISCUSSION RESULTS & DISCUSSION We would like to thank the Hugh and Hazel Darling Foundation for funding this project. We would also like to acknowledge the Office of Undergraduate Research and Creative Scholarship at California Lutheran University. Finally, we thank Dr. Craig Reinhart for his valuable input. ACKNOWLEDGEMENTS ALGORITHMIC FRAMEWORK 1. Dayhoff, J.E., DeLeo, J.M., Artificial Neural Networks Opening the Black Box, Cancer, 2001, 91, 1615-1635. 2. Hanrahan, G Computational Neural Networks Driving Complex Analytical Problem Solving, Anal. Chem, 2010, 82, 4307-4313. 3. Lopez, R., OpenNN: Open Neural Networks Library (Version 0.9) [software]., 2012, Retrieved from http://flood.sourceforge.net. REFERENCES Fig. 3. The architecture of a white box model for ANNs where PSO will be used to train the data. Fig. 4. Pseudo code for ANN framework and logic for the input layer, hidden layer and output layer of the algorithm. In the future, we would like to use ANNs for structural characterization and trend analysis, structure activity relationships, and for instrument optimization. A graphical user interface for the neural network would help students or instructors to easily use and create a neural network based off in their needs. FUTURE WORK Fig. 1. Vector representation of PSO velocity and position updates in a two- dimensional space. At each iteration k a particle updates its velocity and position using Equations (1) and (2), respectively. This process allows all particles in the swarm to update their and . The solution vector (the position vector of the swarm of the) gives the optimal set of values for the PSO parameters s1, s2, and wk . Fig. 2. Visual representation of the architecture of a black box model for ANN. The input layer (representing the data), the hidden layer (where the transformation of our inputs through a function occurs), and the output layer (representing the predictions computed from hidden layer). While the network = not trained for i in each data[i] in dataset y = Σ wi * datai - T a = 1 / ( 1 + e –datai) #activation function if y != target #doesn’t reach target value network = not trained for each wi modify weight_connection(wi) if y = target and reaches end of data set network = trained if y = target #if it reaches the target value repeat for data[i+1] )(111 i k i k i kk i k vprsvwv  i k i k i k vxx 11   Fig. 5. Sample prompt from a run for the ANN algorithm that was written in C++ and implemented in Microsoft Visual Studio.