Package deepnetts.net

Class NeuralNetwork<T extends Trainer>

java.lang.Object

deepnetts.net.NeuralNetwork<T>

All Implemented Interfaces:

TrainerProvider<T>, Serializable

Direct Known Subclasses:

ConvolutionalNetwork, FeedForwardNetwork

public class NeuralNetwork<T extends Trainer>
extends Object
implements TrainerProvider<T>, Serializable

Base class for all neural networks in Deep Netts. Holds a list of abstract layers and loss function. Provides methods for forward and backward calculation, and to access input and output layers. Also provides network and output labels.

See Also:

• AbstractLayer
• LossFunction
• Serialized Form

Method Summary

Modifier and Type	Method	Description
void	applyWeightChanges()	Applies calculated weight changes to all layers.
void	backward()	Performs a backward bass across all layers in neural network, which is the calculation of corrections for the network internal parameters (weights).
InputLayer	getInputLayer()	Returns the input layer of this neural network.
float	getL1RegSum()	Calculates and returns L1 regularization sum of the entire network (all layers included).
float	getL2RegSum()	Calculates and returns L2 regularization sum of the entire network (all layers included).
String	getLabel()	Returns the label(name) of this neural network
AbstractLayer	getLayerAt(int idx)
List<AbstractLayer>	getLayers()	Gets layers of this neural network.
LossFunction	getLossFunction()	Returns a loss function of this network, which is used to calculate total network error during the training.
Mode	getMode()
AbstractScaler	getNormalizer()	Returns data normalization method that is applied to network's inputs.
TensorBase	getOutput()	Returns network's output.
String	getOutputLabel(int i)	Gets a label of the i-th output this network.
String[]	getOutputLabels()	Returns all labels for outputs of this network.
OutputLayer	getOutputLayer()	Returns the output layer of this network.
Preprocessing<TensorBase>	getPreprocessing()	Gets preprocessing that needs to be performed before input is fed to this network.
DeepNettsThreadPool	getThreadPool()
T	getTrainer()	Returns a training algorithm of this neural network.
static <T> T	load(String fileName, Class<T> clazz)	Loads and returns neural network previously saved to a file.
TensorBase	predict(TensorBase input)	Returns the prediction of this neural network for the given input.
void	save(String fileName)	Saves this network using serialization to file with specified fileName.
void	setInput(TensorBase inputs)	Sets network input and calculates entire network (triggers forward pass).
void	setLabel(String label)	Sets label(name) for this neural network.
void	setLossFunction(LossFunction lossFunction)	Sets a loss function of this network, which is used to calculate total network error during the training.
void	setMode(Mode mode)
void	setNormalizer(AbstractScaler normalizer)	Sets normalization data normalization method that is applied to network's inputs.
void	setOutputError(TensorBase outputErrors)	Sets the network's output errors, which are a difference between actual(predicted) and target output.
void	setOutputLabels(String... outputLabels)	Sets output labels of this network.
void	setPreprocessing(Preprocessing<TensorBase> preprocessing)	Sets preprocessing that needs to be performed before input is fed to this network.
void	setTrainer(T trainer)	Sets the training algorithm of this neural network.
javax.visrec.ml.eval.EvaluationMetrics	test(javax.visrec.ml.data.DataSet<? extends MLDataItem> testSet)	Tests how good are predictions of this network using specified test set.
String	toString()	Returns string representation of this network including all layers and settings.
void	train(javax.visrec.ml.data.DataSet<? extends MLDataItem> trainingSet)	Trains the neural network using specified training set.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Method Details

setInput

public void setInput(TensorBase inputs)

Sets network input and calculates entire network (triggers forward pass).

Parameters:

inputs - input tensor

getOutput

public TensorBase getOutput()

Returns network's output.

Returns:

network's output

setOutputError

public void setOutputError(TensorBase outputErrors)

Sets the network's output errors, which are a difference between actual(predicted) and target output.

Parameters:

outputErrors - array of errors for each output, a difference between actual(predicted) and target output

train

public void train(javax.visrec.ml.data.DataSet<? extends MLDataItem> trainingSet)

Trains the neural network using specified training set.

Parameters:

trainingSet - example data given as (input, output) pairs to train the network

predict

public TensorBase predict(TensorBase input)

Returns the prediction of this neural network for the given input. This is the main method to use a trained neural network for inference/prediction. A well trained neural network should provide predictions with low error. Both input and returned prediction are tensors, which are essentially multidimensional arrays.

Parameters:

input - input for the neural network

Returns:

network's prediction.

See Also:

• TensorBase

test

public javax.visrec.ml.eval.EvaluationMetrics test(javax.visrec.ml.data.DataSet<? extends MLDataItem> testSet)

Tests how good are predictions of this network using specified test set. Automatically detects which type of task is network configured to perform and applies appropriate evaluation/test procedure using corresponding Evaluator.

Parameters:

testSet - data set to test/evaluate predictions

Returns:

evaluation metrics that show how good this network is at predicting unseen data

See Also:

• EvaluationMetrics
• ClassificationMetrics
• RegressionMetrics
• Evaluators

applyWeightChanges

public void applyWeightChanges()

Applies calculated weight changes to all layers.

backward

public void backward()

Performs a backward bass across all layers in neural network, which is the calculation of corrections for the network internal parameters (weights). This method invokes the training steps for all the layers starting from the last/output layer and going backwards to first/input layer.

getLayers

public List<AbstractLayer> getLayers()

Gets layers of this neural network.

Returns:

layers of this neural network

getLayerAt

public AbstractLayer getLayerAt(int idx)

getInputLayer

public InputLayer getInputLayer()

Returns the input layer of this neural network. Input layer is the first layer in network which accepts the external input for the network, and forwards it to the next layer in the network.

Returns:

input layer of this neural network

See Also:

• InputLayer

getOutputLayer

public OutputLayer getOutputLayer()

Returns the output layer of this network. Output layer is the last layer of the network which provides final result of the network - predictions.

Returns:

output layer of this network.

See Also:

• OutputLayer

setOutputLabels

public void setOutputLabels(String... outputLabels)

Sets output labels of this network.

Parameters:

outputLabels - labels which correspond to outputs of the network.

getOutputLabels

public String[] getOutputLabels()

Returns all labels for outputs of this network. Each output of the network should have a label which describes what that output represents.

Returns:

labels for outputs of this network.

getOutputLabel

public String getOutputLabel(int i)

Gets a label of the i-th output this network. Each output of the network should have a label which describes what that output represents.

Parameters:

i - idx position of the output

Returns:

label for the i-th output

getLossFunction

public LossFunction getLossFunction()

Returns a loss function of this network, which is used to calculate total network error during the training.

Returns:

loss function of this network

See Also:

• LossFunction
• LossType

setLossFunction

public void setLossFunction(LossFunction lossFunction)

Sets a loss function of this network, which is used to calculate total network error during the training.

Parameters:

lossFunction - loss function to use during the training

getLabel

public String getLabel()

Returns the label(name) of this neural network

Returns:

label of this network

setLabel

public void setLabel(String label)

Sets label(name) for this neural network.

Parameters:

label - label for this network

getL2RegSum

public float getL2RegSum()

Calculates and returns L2 regularization sum of the entire network (all layers included). This value is used during the training to prevent over-fitting.

Returns:

L2 regularization sum

getL1RegSum

public float getL1RegSum()

Calculates and returns L1 regularization sum of the entire network (all layers included). This value is used during the training to prevent over-fitting.

Returns:

L2 regularization sum

getTrainer

public T getTrainer()

Returns a training algorithm of this neural network. Training algorithm performs tuning of the network's internal parameter(weights) in order to minimize an error.

Specified by:

getTrainer in interface TrainerProvider<T extends Trainer>

Returns:

training algorithm of this network

See Also:

• BackpropagationTrainer

setTrainer

public void setTrainer(T trainer)

Sets the training algorithm of this neural network.

Specified by:

setTrainer in interface TrainerProvider<T extends Trainer>

Parameters:

trainer - training algorithm to use for this network

See Also:

• BackpropagationTrainer

getNormalizer

public AbstractScaler getNormalizer()

Returns data normalization method that is applied to network's inputs.

Returns:

setNormalizer

public void setNormalizer(AbstractScaler normalizer)

Sets normalization data normalization method that is applied to network's inputs.

Parameters:

normalizer -

toString

public String toString()

Returns string representation of this network including all layers and settings.

Overrides:

toString in class Object

Returns:

string representation of this network

getPreprocessing

public Preprocessing<TensorBase> getPreprocessing()

Gets preprocessing that needs to be performed before input is fed to this network.

Returns:

setPreprocessing

public void setPreprocessing(Preprocessing<TensorBase> preprocessing)

Sets preprocessing that needs to be performed before input is fed to this network.

Parameters:

preprocessing -

save

public void save(String fileName)
          throws IOException

Saves this network using serialization to file with specified fileName.

Parameters:

fileName - name of the file to save network

Throws:

IOException

load

public static <T> T load(String fileName,
 Class<T> clazz)
                  throws IOException,
ClassNotFoundException

Loads and returns neural network previously saved to a file.

Type Parameters:

T - type(class) of the network to lead and return.

Parameters:

fileName - name of the file to load network from

clazz - class of the neural network to load

Returns:

loaded neural network

Throws:

IOException

ClassNotFoundException

getMode

public Mode getMode()

setMode

public void setMode(Mode mode)

getThreadPool

public DeepNettsThreadPool getThreadPool()