NetTraining Class Reference

#include <NetTraining.h>

Public Member Functions
	NetTraining (ConnectedNet *net, std::vector< TrainingData > trainingData, std::vector< TrainingData > evalData, StopCondition *stopper, double learningRate, int trainingPerIteration=1)
void	run ()

Detailed Description

Training session of a neural net. Consists of a collection of data, net and some stop condition for how long to keep training.

Constructor & Destructor Documentation

NetTraining::NetTraining	(	ConnectedNet *	net,
		std::vector< TrainingData >	trainingData,
		std::vector< TrainingData >	evalData,
		StopCondition *	stopper,
		double	learningRate,
		int	trainingPerIteration = 1
	)

Create a new NetTraining session for the given net with the given data and stop condition

Parameters

net	the neural net to train
trainingData	vector of all data pieces to use for training
evalData	data pieces to use for evaluating performance of the net
stopper	condition for stopping training
learningRate	rate to adjust weights in backpropagation algorithm
trainingPerIteration	times training data should be iterated through before performace is evaluated

Member Function Documentation

void NetTraining::run

(

)

Start the training and run until the stop condition is met

---

The documentation for this class was generated from the following files:

• src/NetTraining.h
• src/NetTraining.cpp