pydbm.cnn.layerablecnn package

Submodules

pydbm.cnn.layerablecnn.convolution_layer module

class pydbm.cnn.layerablecnn.convolution_layer.ConvolutionLayer

Bases: pydbm.cnn.layerable_cnn.LayerableCNN

Convolution Layer.

References

  • Dumoulin, V., & V,kisin, F. (2016). A guide to convolution arithmetic for deep learning. arXiv preprint arXiv:1603.07285.
back_propagate

Back propagation in CNN layers.

Override.

Parameters:delta_arr – 4-rank array like or sparse matrix.
Returns:4-rank array like or sparse matrix.
convolve

Convolution.

Parameters:
  • img_arr – 4-rank array like or sparse matrix.
  • no_bias_flag – Use bias or not.
Returns:

4-rank array like or sparse matrix.
deconvolve

Deconvolution also called transposed convolutions “work by swapping the forward and backward passes of a convolution.” (Dumoulin, V., & Visin, F. 2016, p20.)

Parameters:delta_arr – 4-rank array like or sparse matrix.
Returns:Tuple data. - 4-rank array like or sparse matrix., - 2-rank array like or sparse matrix.

References

  • Dumoulin, V., & V,kisin, F. (2016). A guide to convolution arithmetic for deep learning. arXiv preprint arXiv:1603.07285.
delta_bias_arr

getter

delta_weight_arr

getter

forward_propagate

Forward propagation in CNN layers.

Override.

Parameters:img_arr – 4-rank array like or sparse matrix.
Returns:4-rank array like or sparse matrix.
get_delta_bias_arr

getter

get_delta_weight_arr

getter

get_graph

getter

get_img_arr

getter

get_reshaped_img_arr

getter

get_reshaped_weight_arr

getter

graph

getter

img_arr

getter

reshaped_img_arr

getter

reshaped_weight_arr

getter

set_delta_bias_arr

setter

set_delta_weight_arr

setter

set_img_arr

setter

set_readonly

setter

set_reshaped_img_arr

setter

set_reshaped_weight_arr

setter

pydbm.cnn.layerablecnn.max_pooling_layer module

class pydbm.cnn.layerablecnn.max_pooling_layer.MaxPoolingLayer

Bases: pydbm.cnn.layerable_cnn.LayerableCNN

Max Pooling Layer.

back_propagate

Back propagation in CNN layers.

Override.

Parameters:delta_arr – 4-rank array like or sparse matrix.
Returns:4-rank array like or sparse matrix.
convolve

Convolution.

Parameters:
  • img_arr – 4-rank array like or sparse matrix.
  • no_bias_flag – Use bias or not.
Returns:

4-rank array like or sparse matrix.
deconvolve

Deconvolution also called transposed convolutions “work by swapping the forward and backward passes of a convolution.” (Dumoulin, V., & Visin, F. 2016, p20.)

Parameters:delta_arr – 4-rank array like or sparse matrix.
Returns:Tuple data. - 4-rank array like or sparse matrix., - 2-rank array like or sparse matrix.

References

  • Dumoulin, V., & V,kisin, F. (2016). A guide to convolution arithmetic for deep learning. arXiv preprint arXiv:1603.07285.
delta_bias_arr

getter

delta_weight_arr

getter

forward_propagate

Forward propagation in CNN layers.

Override.

Parameters:matriimg_arr – 4-rank array like or sparse matrix.
Returns:4-rank array like or sparse matrix.
get_delta_bias_arr

getter

get_delta_weight_arr

getter

get_graph

getter

graph

getter

set_delta_bias_arr

setter

set_delta_weight_arr

setter

set_readonly

setter

Module contents