SHOGUN  4.2.0
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Modules Pages
NeuralConvolutionalLayer.h
Go to the documentation of this file.
1 /*
2  * Copyright (c) 2014, Shogun Toolbox Foundation
3  * All rights reserved.
4  *
5  * Redistribution and use in source and binary forms, with or without
6  * modification, are permitted provided that the following conditions are met:
7 
8  * 1. Redistributions of source code must retain the above copyright notice,
9  * this list of conditions and the following disclaimer.
10  *
11  * 2. Redistributions in binary form must reproduce the above copyright notice,
12  * this list of conditions and the following disclaimer in the documentation
13  * and/or other materials provided with the distribution.
14  *
15  * 3. Neither the name of the copyright holder nor the names of its
16  * contributors may be used to endorse or promote products derived from this
17  * software without specific prior written permission.
18 
19  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
20  * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22  * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
23  * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
24  * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
25  * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
26  * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
27  * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
28  * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
29  * POSSIBILITY OF SUCH DAMAGE.
30  *
31  * Written (W) 2014 Khaled Nasr
32  */
33 
34 #ifndef __NEURALCONVOLUTIONALLAYER_H__
35 #define __NEURALCONVOLUTIONALLAYER_H__
36 
37 #include <shogun/lib/common.h>
40 
41 namespace shogun
42 {
43 
47 {
50  NORMAL=0,
57 };
58 
76 {
77 public:
80 
102  int32_t num_maps,
103  int32_t radius_x, int32_t radius_y,
104  int32_t pooling_width=1, int32_t pooling_height=1,
105  int32_t stride_x=1, int32_t stride_y=1,
106  EInitializationMode initialization_mode=NORMAL);
107 
109 
117  virtual void set_batch_size(int32_t batch_size);
118 
128  virtual void initialize_neural_layer(CDynamicObjectArray* layers,
129  SGVector<int32_t> input_indices);
130 
145  virtual void initialize_parameters(SGVector<float64_t> parameters,
146  SGVector<bool> parameter_regularizable,
147  float64_t sigma);
148 
158  virtual void compute_activations(SGVector<float64_t> parameters,
159  CDynamicObjectArray* layers);
160 
186  virtual void compute_gradients(SGVector<float64_t> parameters,
187  SGMatrix<float64_t> targets,
188  CDynamicObjectArray* layers,
189  SGVector<float64_t> parameter_gradients);
190 
198 
207  virtual void enforce_max_norm(SGVector<float64_t> parameters,
208  float64_t max_norm);
209 
210  virtual const char* get_name() const { return "NeuralConvolutionalLayer"; }
211 
212 private:
213  void init();
214 
215 protected:
217  int32_t m_num_maps;
218 
220  int32_t m_input_width;
221 
223  int32_t m_input_height;
224 
227 
229  int32_t m_radius_x;
230 
232  int32_t m_radius_y;
233 
236 
239 
241  int32_t m_stride_x;
242 
244  int32_t m_stride_y;
245 
248 
251 
254 
257 
260 };
261 
262 }
263 #endif
EConvMapActivationFunction m_activation_function
virtual void compute_activations(SGVector< float64_t > parameters, CDynamicObjectArray *layers)
virtual void set_batch_size(int32_t batch_size)
virtual void enforce_max_norm(SGVector< float64_t > parameters, float64_t max_norm)
SGMatrix< float64_t > m_convolution_output_gradients
virtual void compute_gradients(SGVector< float64_t > parameters, SGMatrix< float64_t > targets, CDynamicObjectArray *layers, SGVector< float64_t > parameter_gradients)
Base class for neural network layers.
Definition: NeuralLayer.h:87
double float64_t
Definition: common.h:50
virtual void initialize_parameters(SGVector< float64_t > parameters, SGVector< bool > parameter_regularizable, float64_t sigma)
Dynamic array class for CSGObject pointers that creates an array that can be used like a list or an a...
virtual void initialize_neural_layer(CDynamicObjectArray *layers, SGVector< int32_t > input_indices)
virtual const char * get_name() const
all of classes and functions are contained in the shogun namespace
Definition: class_list.h:18
Main component in convolutional neural networks
virtual float64_t compute_error(SGMatrix< float64_t > targets)
EConvMapActivationFunction
Determines the activation function for neurons in a convolutional feature map.

SHOGUN Machine Learning Toolbox - Documentation