SHOGUN  4.1.0
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Modules Pages
NeuralInputLayer.cpp
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 
35 
36 using namespace shogun;
37 
39 {
40  init();
41 }
42 
43 CNeuralInputLayer::CNeuralInputLayer(int32_t num_neurons, int32_t start_index):
44 CNeuralLayer(num_neurons)
45 {
46  init();
47  m_start_index = start_index;
48 }
49 
50 CNeuralInputLayer::CNeuralInputLayer(int32_t width, int32_t height,
51  int32_t num_channels, int32_t start_index): CNeuralLayer(width*height*num_channels)
52 {
53  init();
54  m_width = width;
55  m_height = height;
56  m_start_index = start_index;
57 }
58 
60 {
61  if (m_start_index == 0)
62  {
63  memcpy(m_activations.matrix, inputs.matrix,
65  }
66  else
67  {
68  for (int32_t i=0; i<m_num_neurons; i++)
69  for (int32_t j=0; j<m_batch_size; j++)
70  m_activations(i,j) = inputs(m_start_index+i, j);
71  }
72  if (gaussian_noise > 0)
73  {
74  int32_t len = m_num_neurons*m_batch_size;
75  for (int32_t k=0; k<len; k++)
77  }
78 }
79 
80 void CNeuralInputLayer::init()
81 {
82  m_start_index = 0;
83  gaussian_noise = 0;
84  SG_ADD(&m_start_index, "start_index",
85  "Start Index", MS_NOT_AVAILABLE);
86  SG_ADD(&gaussian_noise, "gaussian_noise",
87  "Gaussian Noise Standard Deviation", MS_NOT_AVAILABLE);
88 }
static float32_t normal_random(float32_t mean, float32_t std_dev)
Definition: Math.h:1095
SGMatrix< float64_t > m_activations
Definition: NeuralLayer.h:376
Base class for neural network layers.
Definition: NeuralLayer.h:87
double float64_t
Definition: common.h:50
all of classes and functions are contained in the shogun namespace
Definition: class_list.h:18
virtual void compute_activations(SGMatrix< float64_t > inputs)
#define SG_ADD(...)
Definition: SGObject.h:81

SHOGUN Machine Learning Toolbox - Documentation