SHOGUN
v1.1.0
|
00001 /* 00002 * Copyright (c) 2009 Yahoo! Inc. All rights reserved. The copyrights 00003 * embodied in the content of this file are licensed under the BSD 00004 * (revised) open source license. 00005 * 00006 * This program is free software; you can redistribute it and/or modify 00007 * it under the terms of the GNU General Public License as published by 00008 * the Free Software Foundation; either version 3 of the License, or 00009 * (at your option) any later version. 00010 * 00011 * Written (W) 2011 Shashwat Lal Das 00012 * Adaptation of Vowpal Wabbit v5.1. 00013 * Copyright (C) 2011 Berlin Institute of Technology and Max-Planck-Society. 00014 */ 00015 00016 #include <shogun/classifier/vw/learners/VwNonAdaptiveLearner.h> 00017 00018 using namespace shogun; 00019 00020 CVwNonAdaptiveLearner::CVwNonAdaptiveLearner() 00021 : CVwLearner() 00022 { 00023 } 00024 00025 CVwNonAdaptiveLearner::CVwNonAdaptiveLearner(CVwRegressor* regressor, CVwEnvironment* vw_env) 00026 : CVwLearner(regressor, vw_env) 00027 { 00028 } 00029 00030 CVwNonAdaptiveLearner::~CVwNonAdaptiveLearner() 00031 { 00032 } 00033 00034 void CVwNonAdaptiveLearner::train(VwExample* &ex, float32_t update) 00035 { 00036 if (fabs(update) == 0.) 00037 return; 00038 vw_size_t thread_mask = env->thread_mask; 00039 00040 vw_size_t thread_num = 0; 00041 float32_t* weights = reg->weight_vectors[thread_num]; 00042 00043 for (vw_size_t* i = ex->indices.begin; i != ex->indices.end; i++) 00044 { 00045 for (VwFeature* f = ex->atomics[*i].begin; f != ex->atomics[*i].end; f++) 00046 weights[f->weight_index & thread_mask] += update * f->x; 00047 } 00048 00049 for (int32_t k = 0; k < env->pairs.get_num_elements(); k++) 00050 { 00051 char* i = env->pairs.get_element(k); 00052 00053 v_array<VwFeature> temp = ex->atomics[(int32_t)(i[0])]; 00054 temp.begin = ex->atomics[(int32_t)(i[0])].begin; 00055 temp.end = ex->atomics[(int32_t)(i[0])].end; 00056 for (; temp.begin != temp.end; temp.begin++) 00057 quad_update(weights, *temp.begin, ex->atomics[(int32_t)(i[1])], thread_mask, update); 00058 } 00059 } 00060 00061 void CVwNonAdaptiveLearner::quad_update(float32_t* weights, VwFeature& page_feature, v_array<VwFeature> &offer_features, vw_size_t mask, float32_t update) 00062 { 00063 vw_size_t halfhash = quadratic_constant * page_feature.weight_index; 00064 update *= page_feature.x; 00065 for (VwFeature* elem = offer_features.begin; elem != offer_features.end; elem++) 00066 weights[(halfhash + elem->weight_index) & mask] += update * elem->x; 00067 }