安德森鸢尾花卉分类 机器学习算法.zip

# 本python脚本可以应用于 ## 1.SVM ## 2.KMeans ## 算法 # 可以用于机器学习实验： ## 1.基于聚类算法的安德森鸢尾花卉分类 ## 2.基于支持向量机的安德森鸢尾花卉分类 # 可选选项 ## 1.切换实验内容（自适应切换） ## 2.选择交叉验证方案 # 函数功能 ## process.__init__() function: ## usage: ## k_folder : ## if k_folder_test == False ,this option is unavailable,else use cross validation(k-folder),default = 1 ## k_folder_test : ## if this option == True , we will use cross validation(k_folder),else,we will use common training way,default = False ## enhancement_level : ## if data_enhancement == False ,this option is unavailable, else we will add noise on each data,default = 0 ## data_enhancement : ## if this option == False, we will never add noice on data,however this option may cause accuracy reducing,else we will add some noise to the data ## size_train : ## if k_folder_test == True, this option is unavailable,else we will split dataset -> train_data(length = int(len(dataset)*size_train)),default = 0.6 ## size_val : ## if k_folder_test == True, this option is unavailable, else we will split dataset -> val_data(length = int(len(dataset)*size_val)),default = 0.2 ## size_test : ## if k_folder_test == True, this option is unavailable, else we will split dataset -> test_data(length = int(len(dataset)*size_test)),default = 0.2 ## random_set : ## if random_set == True, we will disrupt dataset according to the random_seed, else we will never shuffle the dataset, default = True ## random_seed : ## if random_set == False, this option is unavailable, else we will disrupt dataset according to the random_seed, default = 1 ## draw_fusion_matrix : ## if draw_fusion_matrix == True, we will draw confusion matrix,else we will never draw confusion matrix, default = False ## save_P_R : ## if save_P_R == True, we will save P-R curve by matplotlib.pyplot ,else we will never save P_R curve to your computer,default = False ## runtime_broker : ## if runtime_broker == True, we will stop when the photograph is OK,else we will save the photograph directly,default = False ## save_model : ## if save_model == True, we will dump model to your computer (current path) by joblib,else we will not dump model,default = True ## model_name : ## if save_model == False, this option is unavailable,else we will use this name to save model, default = 'model.pkl' ## kernel : ## SupportVectorMachine Kernel function -> default = 'linear' ## help : ## get_help,default = False ## experiment_ver : ## experiment version you can choose 2 or 3 to train with different type of model (exp_2 : SupportVectorMechine,exp_3 : K-NearestNeighbor) ## n_neighbor : ## define n_neighbor(from sklearn.neighbor.KNeighborsClassifier(n_neighbor)) only experiment_ver == 3,this choice is available # copyright ## this file is edited by lry ## all rights reserved (c) 2020~2023