java.lang.Object
- io.bhagat.ai.supervised.NeuralNetwork

All Implemented Interfaces:

java.io.Serializable
```
public class NeuralNetwork
extends java.lang.Object
implements java.io.Serializable
```
a class for a Neural Network that will take in inputs and send them through difference layers and generate outputs

Author:

Bhagat

See Also:

Serialized Form

Field Summary

Fields
Modifier and Type	Field	Description
`private SerializableFunction<java.lang.Double,java.lang.Double>`	`activationFunction`	the activation function
`private Matrix[]`	`bias`	the bias for each layer
`static SerializableFunction<java.lang.Double,java.lang.Double>`	`defaultActivationFunction`	the default activation function the sigmoid function takes any numbers and puts it between 0 and 1
`private double`	`learingRate`	the learning rate
`private int`	`numOfInputs`	the number of inputs
`private int`	`numOfOutputs`	the number of outputs
`private int[]`	`numsOfHiddens`	the number of hidden nodes in each hidden layer
`private static long`	`serialVersionUID`
`private int[]`	`shape`	an array defining the shape of the network
`private Matrix[]`	`weights`	the weights for each layer

Constructor Summary

Constructors
Constructor	Description
`NeuralNetwork(int... shape)`	Creates a NeuralNetwork with a specified shape
`NeuralNetwork(SerializableFunction<java.lang.Double,java.lang.Double> activationFunction, int... shape)`	Creates a NeuralNetwork with a specified shape

Method Summary

All Methods Instance Methods Concrete Methods
Modifier and Type	Method	Description
`double[]`	`feedForward(double... inputs)`	the feed forward algorithm for making a guess based on the inputs
`void`	`feedForward(DataPoint dataPoint)`	sets the output of feedForward into the dataPoint
`Matrix`	`feedForward(Matrix inputs)`	the feed forward algorithm for making a guess based on the inputs
`Vector`	`feedForward(Vector inputs)`	the feed forward algorithm for making a guess based on the inputs
`SerializableFunction<java.lang.Double,java.lang.Double>`	`getActivationFunction()`
`Matrix[]`	`getBias()`
`double`	`getLearingRate()`
`int`	`getNumOfInputs()`
`int`	`getNumOfOutputs()`
`int[]`	`getNumsOfHiddens()`
`Matrix[]`	`getWeights()`
`private void`	`readObject(java.io.ObjectInputStream in)`
`void`	`serialize(java.lang.String filename)`	Serializes the neural network into a file
`void`	`setActivationFunction(SerializableFunction<java.lang.Double,java.lang.Double> activationFunction)`
`void`	`setBias(Matrix[] bias)`
`void`	`setLearingRate(double learingRate)`
`void`	`setWeights(Matrix[] weights)`
`double`	`test(DataSet dataSet)`	tests the data
`double`	`test(DataSet dataSet, boolean log)`	tests the data
`void`	`train(double[] inputs, double[] targets)`	trains the network based on certain inputs and the known targets for those inputs
`void`	`train(DataPoint dataPoint)`	trains the network using a data point
`void`	`train(DataSet dataSet)`	trains the network using all the data points in the data set
`void`	`train(DataSet dataSet, int epoch)`	trains the network using all the data points in the data set
`void`	`train(Matrix inputs, Matrix targets)`	trains the network based on certain inputs and the known targets for those inputs
`void`	`train(Vector inputs, Vector targets)`	trains the network based on certain inputs and the known targets for those inputs

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

- Field Detail
  - serialVersionUID
```
private static final long serialVersionUID
```
    See Also:
    
    Constant Field Values
  - shape
```
private int[] shape
```
    an array defining the shape of the network
  - numOfInputs
```
private int numOfInputs
```
    the number of inputs
  - numOfOutputs
```
private int numOfOutputs
```
    the number of outputs
  - numsOfHiddens
```
private int[] numsOfHiddens
```
    the number of hidden nodes in each hidden layer
  - weights
```
private Matrix[] weights
```
    the weights for each layer
  - bias
```
private Matrix[] bias
```
    the bias for each layer
  - learingRate
```
private double learingRate
```
    the learning rate
  - activationFunction
```
private SerializableFunction<java.lang.Double,java.lang.Double> activationFunction
```
    the activation function
  - defaultActivationFunction
```
public static SerializableFunction<java.lang.Double,java.lang.Double> defaultActivationFunction
```
    the default activation function the sigmoid function takes any numbers and puts it between 0 and 1
- Constructor Detail
  - NeuralNetwork
```
public NeuralNetwork(SerializableFunction<java.lang.Double,java.lang.Double> activationFunction,
                     int... shape)
```
    Creates a NeuralNetwork with a specified shape
    
    Parameters:
    
    shape - an array defining the shape of the NeuralNetwork
    
    activationFunction - the activation function
  - NeuralNetwork
```
public NeuralNetwork(int... shape)
```
    Creates a NeuralNetwork with a specified shape
    
    Parameters:
    
    shape - an array defining the shape of the NeuralNetwork
- Method Detail
  - feedForward
```
public void feedForward(DataPoint dataPoint)
```
    sets the output of feedForward into the dataPoint
    
    Parameters:
    
    dataPoint - the data point
  - feedForward
```
public double[] feedForward(double... inputs)
```
    the feed forward algorithm for making a guess based on the inputs
    
    Parameters:
    
    inputs - an array of inputs
    
    Returns:
    
    an array for the outputs
  - feedForward
```
public Vector feedForward(Vector inputs)
```
    the feed forward algorithm for making a guess based on the inputs
    
    Parameters:
    
    inputs - a Vector that hold the inputs
    
    Returns:
    
    a Vector for the outputs
  - feedForward
```
public Matrix feedForward(Matrix inputs)
```
    the feed forward algorithm for making a guess based on the inputs
    
    Parameters:
    
    inputs - a Matrix that hold the inputs
    
    Returns:
    
    a Matrix for the outputs
  - train
```
public void train(DataSet dataSet)
```
    trains the network using all the data points in the data set
    
    Parameters:
    
    dataSet - the data set to train the network with
  - train
```
public void train(DataSet dataSet,
                  int epoch)
```
    trains the network using all the data points in the data set
    
    Parameters:
    
    dataSet - the data set to train the network with
    
    epoch - how many times to train each data point
  - train
```
public void train(DataPoint dataPoint)
```
    trains the network using a data point
    
    Parameters:
    
    dataPoint - the data point
  - train
```
public void train(double[] inputs,
                  double[] targets)
```
    trains the network based on certain inputs and the known targets for those inputs
    
    Parameters:
    
    inputs - the inputs
    
    targets - the targets
  - train
```
public void train(Vector inputs,
                  Vector targets)
```
    trains the network based on certain inputs and the known targets for those inputs
    
    Parameters:
    
    inputs - the inputs
    
    targets - the targets
  - train
```
public void train(Matrix inputs,
                  Matrix targets)
```
    trains the network based on certain inputs and the known targets for those inputs
    
    Parameters:
    
    inputs - the inputs
    
    targets - the targets
  - test
```
public double test(DataSet dataSet)
```
    tests the data
    
    Parameters:
    
    dataSet - the testing data set
    
    Returns:
    
    the accuracy
  - test
```
public double test(DataSet dataSet,
                   boolean log)
```
    tests the data
    
    Parameters:
    
    dataSet - the testing data set
    
    log - whether or not to log out the tests
    
    Returns:
    
    the accuracy
  - readObject
```
private void readObject(java.io.ObjectInputStream in)
                 throws java.io.IOException,
                        java.lang.ClassNotFoundException
```
    Throws:
    
    java.io.IOException
    
    java.lang.ClassNotFoundException
  - serialize
```
public void serialize(java.lang.String filename)
```
    Serializes the neural network into a file
    
    Parameters:
    
    filename - the file name
  - getWeights
```
public Matrix[] getWeights()
```
    Returns:
    
    the weights
  - setWeights
```
public void setWeights(Matrix[] weights)
```
    Parameters:
    
    weights - the weights to set
  - getBias
```
public Matrix[] getBias()
```
    Returns:
    
    the bias
  - setBias
```
public void setBias(Matrix[] bias)
```
    Parameters:
    
    bias - the bias to set
  - getNumOfInputs
```
public int getNumOfInputs()
```
    Returns:
    
    the numOfInputs
  - getNumOfOutputs
```
public int getNumOfOutputs()
```
    Returns:
    
    the numOfOutputs
  - getNumsOfHiddens
```
public int[] getNumsOfHiddens()
```
    Returns:
    
    the numsOfHiddens
  - getActivationFunction
```
public SerializableFunction<java.lang.Double,java.lang.Double> getActivationFunction()
```
    Returns:
    
    the activationFunction
  - setActivationFunction
```
public void setActivationFunction(SerializableFunction<java.lang.Double,java.lang.Double> activationFunction)
```
    Parameters:
    
    activationFunction - the activationFunction to set
  - getLearingRate
```
public double getLearingRate()
```
    Returns:
    
    the learingRate
  - setLearingRate
```
public void setLearingRate(double learingRate)
```
    Parameters:
    
    learingRate - the learingRate to set

Class NeuralNetwork

Field Summary

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Field Detail

serialVersionUID

shape

numOfInputs

numOfOutputs

numsOfHiddens

weights

bias

learingRate

activationFunction

defaultActivationFunction

Constructor Detail

NeuralNetwork

NeuralNetwork

Method Detail

feedForward

feedForward

feedForward

feedForward

train

train

train

train

train

train

test

test

readObject

serialize

getWeights

setWeights

getBias

setBias

getNumOfInputs

getNumOfOutputs

getNumsOfHiddens

getActivationFunction

setActivationFunction

getLearingRate

setLearingRate