python中实现蚁群算法进行特征选择+SVM分类

import numpy as np from feature_select import Fitness from scipy.io import loadmat from sklearn import preprocessing from sklearn.model_selection import train_test_split input_image=loadmat('C:\\Users\\Huff\\Desktop\\yanxm16\\paper_data\\Indian_pines_corrected.mat')['indian_pines_corrected'] output_image=loadmat('C:\\Users\\Huff\\Desktop\\yanxm16\\paper_data\\Indian_pines_gt.mat')['indian_pines_gt'] new_datawithlabel_list = [] for i in range(output_image.shape[0]): for j in range(output_image.shape[1]): if output_image[i][j] != 0: c2l = list(input_image[i][j]) c2l.append(output_image[i][j]) new_datawithlabel_list.append(c2l) new_datawithlabel_array = np.array(new_datawithlabel_list) data_D = preprocessing.StandardScaler().fit_transform(new_datawithlabel_array[:,:-1]) data_L = new_datawithlabel_array[:,-1] features_num=200 status_num=2 class Graph(object): def __init__(self, features_num: int,status_num:int): self.features_num=features_num self.status_num=status_num self.pheromone = [[1 / (features_num/40) for j in range(features_num)] for i in range(status_num)] class Feature_select_from_index(object): def __init__(self,features_index_set,data_D,data_L): self.features_index_set=features_index_set self.data_D=data_D self.data_L=data_L def feature_select(self): data_after_feature_selected=[] for i in self.features_index_set: data_after_feature_selected.append(self.data_D[:,i]) data_after_feature_selected = np.array(data_after_feature_selected) data_after_feature_selected = data_after_feature_selected.transpose() return data_after_feature_selected def split_selected_features(self): data_after_feature_selected=self.feature_select() return train_test_split(data_after_feature_selected, self.data_L, test_size=0.3, random_state=34) class ACO(object): def __init__(self, ant_count: int, generations: int, alpha: float, beta: float, rho: float, q: int, strategy: int): self.Q = q self.rho = rho self.beta = beta self.alpha = alpha self.ant_count = ant_count self.generations = generations self.update_strategy = strategy self.ant = _Ant def _update_pheromone(self, graph: Graph, ants: list): for i, row in enumerate(graph.pheromone): for j, col in enumerate(row): graph.pheromone[i][j] *= self.rho for ant in ants: graph.pheromone[i][j] += ant.pheromone_delta[i][j] def _update_pheromone_init(self, graph: Graph,ant): for i, row in enumerate(graph.pheromone): for j, col in enumerate(row): graph.pheromone[i][j] *= self.rho graph.pheromone[i][j] += ant.pheromone_delta[i][j] def solve1(self, graph: Graph): best_classification_accuracy = 0 best_solution = [] gamma = 0.125 c = 20 for gen in range(self.generations): print('第{}代蚂蚁开始选择特征'.format(gen) + '。' * 80) ants = [_Ant(self, graph,p0=0.2) for i in range(self.ant_count)] # 生成ant_count个_Ant对象 for ant in ants: if gen == 0 and ants.index(ant) == 0: print('第0代第0只蚂蚁进行特征选择初始化......') features_index_set, features_status = ant.init_feature_set_and_setStatus() print('第0代第0只蚂蚁选择的特征：features_index_set={},features_status={}'.format(features_index_set, features_status)) print('第0代第0只蚂蚁选择的特征:len(features_index_set)={},len(features_status)={}'.format( len(features_index_set), len(features_status))) feature_selected_object = Feature_select_from_index(features_index_set, data_D, data_L) train, test, train_result, test_result = feature_selected_object.split_selected_features() # print (train.shape,test.shape) best_classification_accuracy = Fitness.funcFitness(train, train_result, test, test_result, gamma, c) print('第0代第0只蚂蚁选择的特征分类的精确度为：{}'.format(best_classification_accuracy)) ant._update_pheromone_delta(features_status) self._update_pheromone_init(graph, ant) else: print('第{}代第{}只蚂蚁开始选择'.format(gen, ants.index(ant))) for j in range(graph.features_num): ant._select_next_feature_status1(j) features_index_set = ant.features_index_set features_status = ant.features_status print('len(features_index_set)={},len(features_status)={}'.format(len(features_index_set), len(features_status))) features = Feature_select_from_index(features_index_set, data_D, data_L) train, test, train_result, test_result = features.split_selected_features() fitness = Fitness.funcFitness(train, train_result, test, test_result, gamma, c) if fitness > best_classification_accuracy: print('本次分类准确率优于上次' + '*' * 30+'acc={}'.format(fitness)) best_classification_accuracy = fitness best_solution = [] + ant.features_status # update pheromone ant._update_pheromone_delta(features_status) self._update_pheromone(graph, ants) return best_solution, best_classification_accuracy class _Ant(object): def __init__(self, aco: ACO, graph: Graph,p0:0.2): self.colony = aco self.graph = graph self.pheromone_delta = [] # the local increase of pheromone self.allowed_features_index=[i+1 for i in range(graph.features_num)] self.features_index_set=[] self.features_status = [] self.p0=p0 def init_feature_set_and_setStatus(self): init_feature_status = list(np.random.rand(features_num)) for j in range(self.graph.features_num): if init_feature_status[j]<self.p0: self.features_index_set.append(j) self.features_status.append(1) else: self.features_status.append(0) return self.features_index_set,self.features_status def _select_next_feature_status1(self,j): # print('选择第{}个波段开始'.format(j)) init_feature_status = np.random.rand() if init_feature_status<=self.p0: pheromone_concentration=[] for i in range(self.graph.status_num): pheromone_concentration.append(self.graph.pheromone[i][j]) status=pheromone_concentration.index(max(pheromone_concentration)) self.features_status.append(status) if status==1: self.features_index_set.append(j) else: denominator=0 pheromone_concentration_scaled = [] for i in range(self.graph.status_num): denominator+=self.graph.pheromone[i][j] for i in range(self.graph.status_num): self.graph.pheromone[i][j]=self.graph.pheromone[i][j]/denominator pheromone_concentration_scaled.append(self.graph.pheromone[i][j]) init_feature_status_again = np.random.rand() if ini