#include <ebl_layers.h>

Inheritance diagram for ebl::convolution_layer< T, Tstate >:

Public Member Functions
	convolution_layer (parameter< T, Tstate > p, idxdim &ker, idxdim &stride, idx< intg > &tbl, bool tanh=true, const char name="convolution_layer")
void	fprop (Tstate &in, Tstate &out)
	fprop from in to out
void	bprop (Tstate &in, Tstate &out)
	bprop
void	bbprop (Tstate &in, Tstate &out)
	bbprop
void	forget (forget_param_linear &fp)
	initialize the weights to random values
virtual fidxdim	fprop_size (fidxdim &i_size)
virtual fidxdim	bprop_size (const fidxdim &o_size)
virtual convolution_layer< T, Tstate > *	copy ()
	Returns a deep copy of this module.
Public Attributes
convolution_module< T, Tstate >	convol
	convolution module
addc_module< T, Tstate >	adder
	bias vector
module_1_1< T, Tstate > *	sigmoid
	non-linear funct
Tstate *	sum
	convolution result

Detailed Description

a convolution neural net layer: convolution + tanh non-linearity.

Constructor & Destructor Documentation

template<typename T , class Tstate >

ebl::convolution_layer< T, Tstate >::convolution_layer	(	parameter< T, Tstate > *	p,
		idxdim &	ker,
		idxdim &	stride,
		idx< intg > &	tbl,
		bool	tanh = `true`,
		const char *	name = `"convolution_layer< T, Tstate >"`
	)

constructor. Arguments are a pointer to a parameter in which the trainable weights will be appended, the number of inputs, and the number of outputs.

Parameters:

p	is used to store all parametric variables in a single place.
ker	The convolution kernel sizes.
stride	The convolution strides.
table	is the convolution connection table.
tanh	If true, use tanh squasher, stdsigmoid otherwise.

template<typename T , class Tstate >

fidxdim ebl::convolution_layer< T, Tstate >::bprop_size ( const fidxdim & o_size ) [virtual]

Return dimensions compatible with this module given output dimensions. See module_1_1_gen's documentation for more details.

template<typename T , class Tstate >

fidxdim ebl::convolution_layer< T, Tstate >::fprop_size ( fidxdim & i_size ) [virtual]

Return dimensions that are compatible with this module. See module_1_1_gen's documentation for more details.

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