Interface Summary Interface Description RandomizerDefines the interface for a class that is capable of randomizing the weights and bias values of a neural network. Class Summary Class Description BasicRandomizerProvides basic functionality that most randomizers will need. ConsistentRandomizerA randomizer that takes a seed and will always produce consistent results. ConstRandomizerA randomizer that will create always set the random number to a const value, used mainly for testing. DistortA randomizer that distorts what is already present in the neural network. FanInRandomizerA randomizer that attempts to create starting weight values that are conducive to propagation training. GaussianRandomizerGenerally, you will not want to use this randomizer as a pure neural network randomizer. NguyenWidrowRandomizerImplementation of Nguyen-Widrow weight initialization. RangeRandomizerA randomizer that will create random weight and bias values that are between a specified range.
SCaVis 2.1 © jWork.ORG