SHOGUN  v3.2.0
WDSVMOcas.h
浏览该文件的文档.
1 /*
2  * This program is free software; you can redistribute it and/or modify
3  * it under the terms of the GNU General Public License as published by
4  * the Free Software Foundation; either version 3 of the License, or
5  * (at your option) any later version.
6  *
7  * Written (W) 2007-2008 Vojtech Franc
8  * Written (W) 2007-2009 Soeren Sonnenburg
9  * Copyright (C) 2007-2009 Fraunhofer Institute FIRST and Max-Planck-Society
10  */
11 
12 #ifndef _WDSVMOCAS_H___
13 #define _WDSVMOCAS_H___
14 
15 #include <shogun/lib/common.h>
16 #include <shogun/machine/Machine.h>
19 #include <shogun/labels/Labels.h>
20 
21 namespace shogun
22 {
23 template <class ST> class CStringFeatures;
24 
26 class CWDSVMOcas : public CMachine
27 {
28  public:
31 
33  CWDSVMOcas();
34 
39  CWDSVMOcas(E_SVM_TYPE type);
40 
49  CWDSVMOcas(
50  float64_t C, int32_t d, int32_t from_d,
51  CStringFeatures<uint8_t>* traindat, CLabels* trainlab);
52  virtual ~CWDSVMOcas();
53 
59 
66  inline void set_C(float64_t c_neg, float64_t c_pos) { C1=c_neg; C2=c_pos; }
67 
72  inline float64_t get_C1() { return C1; }
73 
78  inline float64_t get_C2() { return C2; }
79 
84  inline void set_epsilon(float64_t eps) { epsilon=eps; }
85 
90  inline float64_t get_epsilon() { return epsilon; }
91 
97  {
98  SG_REF(feat);
100  features=feat;
101  }
102 
108  {
109  SG_REF(features);
110  return features;
111  }
112 
117  inline void set_bias_enabled(bool enable_bias) { use_bias=enable_bias; }
118 
123  inline bool get_bias_enabled() { return use_bias; }
124 
129  inline void set_bufsize(int32_t sz) { bufsize=sz; }
130 
135  inline int32_t get_bufsize() { return bufsize; }
136 
142  inline void set_degree(int32_t d, int32_t from_d)
143  {
144  degree=d;
145  from_degree=from_d;
146  }
147 
152  inline int32_t get_degree() { return degree; }
153 
160  virtual CBinaryLabels* apply_binary(CFeatures* data=NULL);
161 
168  virtual CRegressionLabels* apply_regression(CFeatures* data=NULL);
169 
175  virtual float64_t apply_one(int32_t num)
176  {
178  if (!wd_weights)
179  set_wd_weights();
180 
181  int32_t len=0;
182  float64_t sum=0;
183  bool free_vec;
184  uint8_t* vec=features->get_feature_vector(num, len, free_vec);
185  //SG_INFO("len %d, string_length %d\n", len, string_length)
186  ASSERT(len==string_length)
187 
188  for (int32_t j=0; j<string_length; j++)
189  {
190  int32_t offs=w_dim_single_char*j;
191  int32_t val=0;
192  for (int32_t k=0; (j+k<string_length) && (k<degree); k++)
193  {
194  val=val*alphabet_size + vec[j+k];
195  sum+=wd_weights[k] * w[offs+val];
196  offs+=w_offsets[k];
197  }
198  }
199  features->free_feature_vector(vec, num, free_vec);
200  return sum/normalization_const;
201  }
202 
205  {
208  for (int32_t i=0; i<degree; i++)
210 
212  SG_DEBUG("normalization_const:%f\n", normalization_const)
213  }
214 
220 
221 
222  protected:
223 
229 
234  int32_t set_wd_weights();
235 
244  static void compute_W(
245  float64_t *sq_norm_W, float64_t *dp_WoldW, float64_t *alpha,
246  uint32_t nSel, void* ptr );
247 
254  static float64_t update_W(float64_t t, void* ptr );
255 
261  static void* add_new_cut_helper(void* ptr);
262 
271  static int add_new_cut(
272  float64_t *new_col_H, uint32_t *new_cut, uint32_t cut_length,
273  uint32_t nSel, void* ptr );
274 
280  static void* compute_output_helper(void* ptr);
281 
287  static int compute_output( float64_t *output, void* ptr );
288 
295  static int sort( float64_t* vals, float64_t* data, uint32_t size);
296 
298  static inline void print(ocas_return_value_T value)
299  {
300  return;
301  }
302 
303 
305  virtual const char* get_name() const { return "WDSVMOcas"; }
306 
307  protected:
316  virtual bool train_machine(CFeatures* data=NULL);
317 
318  protected:
322  bool use_bias;
324  int32_t bufsize;
332  E_SVM_TYPE method;
333 
335  int32_t degree;
337  int32_t from_degree;
341  int32_t num_vec;
343  int32_t string_length;
345  int32_t alphabet_size;
346 
349 
355  int32_t* w_offsets;
357  int32_t w_dim;
366 
371 };
372 }
373 #endif
float64_t get_C2()
Definition: WDSVMOcas.h:78
EMachineType
Definition: Machine.h:33
float32_t * w
Definition: WDSVMOcas.h:361
virtual EMachineType get_classifier_type()
Definition: WDSVMOcas.h:58
Real Labels are real-valued labels
class WDSVMOcas
Definition: WDSVMOcas.h:26
float32_t * wd_weights
Definition: WDSVMOcas.h:339
float64_t * cp_bias
Definition: WDSVMOcas.h:370
virtual const char * get_name() const
Definition: WDSVMOcas.h:305
static int sort(float64_t *vals, float64_t *data, uint32_t size)
Definition: WDSVMOcas.cpp:439
The class Labels models labels, i.e. class assignments of objects.
Definition: Labels.h:35
static int add_new_cut(float64_t *new_col_H, uint32_t *new_cut, uint32_t cut_length, uint32_t nSel, void *ptr)
Definition: WDSVMOcas.cpp:352
virtual CRegressionLabels * apply_regression(CFeatures *data=NULL)
Definition: WDSVMOcas.cpp:106
static void * add_new_cut_helper(void *ptr)
Definition: WDSVMOcas.cpp:300
#define SG_UNREF(x)
Definition: SGRefObject.h:35
void set_bias_enabled(bool enable_bias)
Definition: WDSVMOcas.h:117
float64_t * lab
Definition: WDSVMOcas.h:365
E_SVM_TYPE method
Definition: WDSVMOcas.h:332
int32_t set_wd_weights()
Definition: WDSVMOcas.cpp:144
int32_t get_bufsize()
Definition: WDSVMOcas.h:135
static int compute_output(float64_t *output, void *ptr)
Definition: WDSVMOcas.cpp:542
int32_t alphabet_size
Definition: WDSVMOcas.h:345
float64_t get_C1()
Definition: WDSVMOcas.h:72
int32_t from_degree
Definition: WDSVMOcas.h:337
float64_t get_normalization_const()
Definition: WDSVMOcas.h:219
bool get_bias_enabled()
Definition: WDSVMOcas.h:123
float64_t epsilon
Definition: WDSVMOcas.h:330
virtual ~CWDSVMOcas()
Definition: WDSVMOcas.cpp:96
A generic learning machine interface.
Definition: Machine.h:138
SGVector< float64_t > apply_get_outputs(CFeatures *data)
Definition: WDSVMOcas.cpp:112
static float64_t update_W(float64_t t, void *ptr)
Definition: WDSVMOcas.cpp:267
virtual CBinaryLabels * apply_binary(CFeatures *data=NULL)
Definition: WDSVMOcas.cpp:100
CStringFeatures< uint8_t > * get_features()
Definition: WDSVMOcas.h:107
#define ASSERT(x)
Definition: SGIO.h:203
Template class StringFeatures implements a list of strings.
Definition: WDSVMOcas.h:23
static void * compute_output_helper(void *ptr)
Definition: WDSVMOcas.cpp:450
CStringFeatures< uint8_t > * features
Definition: WDSVMOcas.h:320
float64_t normalization_const
Definition: WDSVMOcas.h:348
float64_t bias
Definition: WDSVMOcas.h:351
double float64_t
Definition: common.h:48
float64_t old_bias
Definition: WDSVMOcas.h:353
void set_normalization_const()
Definition: WDSVMOcas.h:204
#define SG_REF(x)
Definition: SGRefObject.h:34
void free_feature_vector(ST *feat_vec, int32_t num, bool dofree)
void set_degree(int32_t d, int32_t from_d)
Definition: WDSVMOcas.h:142
void set_epsilon(float64_t eps)
Definition: WDSVMOcas.h:84
virtual bool train_machine(CFeatures *data=NULL)
Definition: WDSVMOcas.cpp:162
static void compute_W(float64_t *sq_norm_W, float64_t *dp_WoldW, float64_t *alpha, uint32_t nSel, void *ptr)
Definition: WDSVMOcas.cpp:612
SGVector< ST > get_feature_vector(int32_t num)
virtual float64_t apply_one(int32_t num)
Definition: WDSVMOcas.h:175
MACHINE_PROBLEM_TYPE(PT_BINARY)
float float32_t
Definition: common.h:47
void set_bufsize(int32_t sz)
Definition: WDSVMOcas.h:129
float32_t ** cuts
Definition: WDSVMOcas.h:368
void set_C(float64_t c_neg, float64_t c_pos)
Definition: WDSVMOcas.h:66
#define SG_DEBUG(...)
Definition: SGIO.h:109
all of classes and functions are contained in the shogun namespace
Definition: class_list.h:16
int32_t get_degree()
Definition: WDSVMOcas.h:152
The class Features is the base class of all feature objects.
Definition: Features.h:62
int32_t * w_offsets
Definition: WDSVMOcas.h:355
int32_t string_length
Definition: WDSVMOcas.h:343
Binary Labels for binary classification
Definition: BinaryLabels.h:36
float32_t * old_w
Definition: WDSVMOcas.h:363
static void print(ocas_return_value_T value)
Definition: WDSVMOcas.h:298
static float32_t sqrt(float32_t x)
x^0.5
Definition: Math.h:245
int32_t w_dim_single_char
Definition: WDSVMOcas.h:359
float64_t get_epsilon()
Definition: WDSVMOcas.h:90
void set_features(CStringFeatures< uint8_t > *feat)
Definition: WDSVMOcas.h:96

SHOGUN Machine Learning Toolbox - Documentation