Namespace NeuralNetworks.Operations.Parameterized
Classes
- BiasAdd
Computes bias addition.
- Conv2D
Dimensions of the input are: [batch, channels, height, width]. Dimensions of the param array are: [channels, filters, kernelSize, kernelSize]. CalcOutput returns an array of dimensions: [batch, filters, height, width]. Padding is assumed to be the same on all sides = kernelSize / 2
- ParamOperation<TIn, TOut, TParam>
An Operation with parameters of type TParam.
- WeightMultiply
Weight multiplication operation for a neural network.