机器学习工作流

 本文的目的是演示如何构建一个相对完整的机器学习工作流

1.首先对工程进行基本的参数配置

# 进行建模基本配置
SCORE_EVA = 'roc_auc'
random_state_clf = 1
n_jobs= 4
cv_split = StratifiedKFold(n_splits=5, shuffle=True, random_state=1)
cv_split2 = StratifiedKFold(n_splits=5, shuffle=True, random_state=2)
X, y = data_of_features, label

2.基于各ML模型默认参数进行建模评估，了解模型针对当前任务的基本建模能力，筛选出有前途的模型进行超参数调优

#Machine Learning Algorithm (MLA) Selection and Initialization
MLA = [#Ensemble Methodsensemble.AdaBoostClassifier(random_state=random_state_clf),ensemble.BaggingClassifier(random_state=random_state_clf),ensemble.ExtraTreesClassifier(random_state=random_state_clf),ensemble.GradientBoostingClassifier(random_state=random_state_clf),ensemble.RandomForestClassifier(random_state=random_state_clf),#Gaussian Processesgaussian_process.GaussianProcessClassifier(random_state=random_state_clf),#GLMlinear_model.LogisticRegressionCV(random_state=random_state_clf),linear_model.PassiveAggressiveClassifier(random_state=random_state_clf),linear_model.RidgeClassifierCV(random_state=random_state_clf),linear_model.SGDClassifier(random_state=random_state_clf),linear_model.Perceptron(random_state=random_state_clf),#Navies Bayesnaive_bayes.BernoulliNB(random_state=random_state_clf),naive_bayes.GaussianNB(random_state=random_state_clf),#Nearest Neighborneighbors.KNeighborsClassifier(random_state=random_state_clf),#SVMsvm.SVC(probability=True,random_state=random_state_clf),svm.NuSVC(probability=True,random_state=random_state_clf),svm.LinearSVC(random_state=random_state_clf),#Trees    tree.DecisionTreeClassifier(random_state=random_state_clf),tree.ExtraTreeClassifier(random_state=random_state_clf),#Discriminant Analysisdiscriminant_analysis.LinearDiscriminantAnalysis(random_state=random_state_clf),discriminant_analysis.QuadraticDiscriminantAnalysis(random_state=random_state_clf),XGBClassifier(random_state=random_state_clf)    ]MLA_columns = ['MLA Name', 'MLA Parameters','MLA Train Matric Mean', 'MLA Test Matric Mean', 'MLA Test Matric 3*STD' ,'MLA Time']
MLA_compare = pd.DataFrame(columns = MLA_columns)
row_index = 0
for alg in MLA:#set name and parametersMLA_name = alg.__class__.__name__MLA_compare.loc[row_index, 'MLA Name'] = MLA_nameMLA_compare.loc[row_index, 'MLA Parameters'] = str(alg.get_params())#score model with cross validation: http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.cross_validate.html#sklearn.model_selection.cross_validatecv_results = model_selection.cross_validate(alg, X, y, cv  = cv_split,return_train_score=True,scoring = SCORE_EVA)MLA_compare.loc[row_index, 'MLA Time'] = cv_results['fit_time'].mean()MLA_compare.loc[row_index, 'MLA Train Matric Mean'] = cv_results['train_score'].mean()MLA_compare.loc[row_index, 'MLA Test Matric Mean'] = cv_results['test_score'].mean()   #if this is a non-bias random sample, then +/-3 standard deviations (std) from the mean, should statistically capture 99.7% of the subsetsMLA_compare.loc[row_index, 'MLA Test Matric 3*STD'] = cv_results['test_score'].std()*3   #let's know the worst that can happen!row_index+=1MLA_compare = MLA_compare.sort_values(by='MLA Test Matric Mean', ascending=False)  # 降序排列
MLA_compare

3.挑选出表现较好的模型，结合交叉验证和递归特征消除技术，同时进行超参数调优和特征选择

#函数可以选择使用REF或者SelectKBest方法结合贝叶斯或网格进行交叉验证寻优
def SearchCV_Feature_and_Parameter(X,y,clf_model,param_grid,cv_split,SCORE_EVA='roc_auc',Search_method ='Bayes',feature_method = 'ref', bayes_n_iter=10,verbose = 0,n_jobs=1):if feature_method == 'ref':pipe = Pipeline( [('scaler', StandardScaler()),('feature_selector', RFE(estimator=clf_model)),('model', clf_model)])else:pipe = Pipeline( [('scaler', StandardScaler()),('feature_selector', SelectKBest(f_classif)),  ('model', clf_model)])if Search_method =='grid':grid_search = GridSearchCV(pipe, param_grid=param_grid, cv=cv_split, verbose=verbose,scoring = SCORE_EVA,n_jobs=n_jobs)else:grid_search = BayesSearchCV(pipe, search_spaces=param_grid, verbose=verbose, scoring=SCORE_EVA, cv=cv_split, n_iter=bayes_n_iter,n_jobs=n_jobs)grid_search.fit(X, y)return grid_search
#函数通过多次，多折交叉验证的形式找到最佳特征集和超参数，在多次的交叉验证结果中，以出现最多的特征和参数作为最终的优选结果。这里如果用贝叶斯方法，可能每次交叉验证的超参数结果都是唯一的，导致所有
#结果出现次数都是1def mutil_times_SearchCV_Feature_and_Parameter(X,y,clf_model,param_grid,cv_outter=10,cv_inner=5,SCORE_EVA='roc_auc',Search_method ='Bayes',feature_method = 'ref',bayes_n_iter=2,verbose=0,n_jobs=1):start_time = timeit.default_timer()inner_cv = StratifiedKFold(n_splits=cv_inner, shuffle=True, random_state=1)if cv_outter==1:grid_search_result = SearchCV_Feature_and_Parameter(X,y,clf_model,param_grid,inner_cv,SCORE_EVA,Search_method,feature_method,bayes_n_iter,verbose,n_jobs)end_time = timeit.default_timer()print(f"函数运行时间为 {(end_time - start_time)/60} 分")print("Best score found: ", grid_search_result.best_score_)print("Best parameters found: ", grid_search_result.best_params_)print("Selected features:", np.array(features43)[grid_search_result.best_estimator_.named_steps['feature_selector'].support_])return grid_search_resultelse:      outer_cv = StratifiedKFold(n_splits=cv_split, shuffle=True, random_state=0)roc =[]best_params_history = [] selected_features_history = [] # 执行超参数优化for train_idx, test_idx in outer_cv.split(X, y):X_train, X_test = X.iloc[train_idx], X.iloc[test_idx]y_train, y_test = y.iloc[train_idx], y.iloc[test_idx]search = SearchCV_Feature_and_Parameter(X_train,y_train,clf_model,param_grid,inner_cv,SCORE_EVA,Search_method,feature_method,bayes_n_iter,verbose,n_jobs)best_params_history.append(search.best_params_)best_model = search.best_estimator_selected_features = best_model.named_steps['feature_selector'].get_support()selected_features_history.append(selected_features)y_pred = best_model.predict(X_test)y_pred_proba = best_model.predict_proba(X_test)[:,1]roc.append(roc_auc_score(y_test, y_pred_proba))best_params_history_df = pd.DataFrame([dict(ordered_dict) for ordered_dict in best_params_history])best_params_history_df[SCORE_EVA] = rocprint(f"{cv_split}次{cv_split}折交叉验证平均ROC: {np.mean(roc):.4f}  std: {np.std(roc):.4f} ,{[round(meta,3) for meta in roc]}")#for i, selected_features in enumerate(selected_features_history, start=1):#    print(f"第{i}次交叉验证所选择的特征： {np.array(features)[selected_features]}")param_names = best_params_history[0].keys()overall_best_params = {}for param_name in param_names:value_counts = Counter([params[param_name] for params in best_params_history])most_common_value = value_counts.most_common(1)[0][0]overall_best_params[param_name] = most_common_valueprint("整体最佳超参数: ", overall_best_params)# 多模型集成+计算整体最佳作为最终超参数total_features = X.shape[1]feature_selection_counts = np.zeros(total_features)# 统计每个特征被选中的次数for selected_features in selected_features_history:feature_selection_counts += selected_features.astype(int)# 设置阈值以确定整体最佳特征集threshold = len(selected_features_history) // 2overall_best_features = feature_selection_counts > thresholdprint("整体最佳特征集: ", np.array(features)[overall_best_features])end_time = timeit.default_timer()print(f"函数运行时间为 {(end_time - start_time)/60} 分")return best_params_history_df,selected_features_history,roc
def model_evaluate(X,y,model,n_times,test_size=0.3):scores = []# 进行多次随机数据划分for i in range(n_times):# 划分训练集和测试集X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=test_size,random_state=i)# 训练并评估模型，每次调用fit都会重新初始化模型权重model.fit(X_train, y_train)y_pred = model.predict_proba(X_test)[:,1]score = roc_auc_score(y_test, y_pred)scores.append(score)return scores
# 通过建立ML工作流对，针对较优模型进行RFE特征选择，和超参数调优，作为本次建模的核心模型
grid_n_estimator = Integer(1, 300)
grid_ratio = Real(0.01, 1.0, 'log-uniform')
grid_learn = Real(0.01, 1.0, 'log-uniform')
grid_max_depth =  Integer(1, 15)
grid_min_samples = [5, 10, .03, .05, .10]
grid_criterion = ['gini', 'entropy']
grid_bool = [True, False]
grid_seed = [0]# 定义超参数搜索空间
param_grid = {#'feature_selector__k': Integer(5, 15),'feature_selector__n_features_to_select': Integer(5, 15),'model__learning_rate': Real(0.01, 1.0, 'log-uniform'),'model__max_depth': Integer(1, 50),'model__n_estimators': Integer(50, 200),'model__random_state': grid_seed
}clf_model = XGBClassifier(scale_pos_weight=2,objective='binary:logistic',seed=0)
grid_search_result = mutil_times_SearchCV_Feature_and_Parameter(X,y,clf_model,param_grid,cv_outter=1,cv_inner=5,SCORE_EVA='roc_auc',Search_method ='Bayes',feature_method = 'ref',bayes_n_iter=10,verbose=0,n_jobs=n_jobs)
# 得到optimal特征子集和超参数后，通过多次数据划分，评估模型整体和泛化性能，其中，泛化性能以std结果体现
X_best = X[np.array(features43)[grid_search_result.best_estimator_.named_steps['feature_selector'].support_]]
X_best = StandardScaler().fit_transform(X_best)
clf_model.set_params(**{k.replace("model__", ""): v for k, v in grid_search_result.best_params_.items() if k.startswith("model__")})
scores = model_evaluate(X_best,y,clf_model_EVA,n_times=100,test_size=0.3)
mean_score = round(np.mean(scores),3)
std_score = round(np.std(scores),3)
print('最佳模型',mean_score,std_score)

4.完成特征选择后，基于优选特征子集，结合贝叶斯交叉验证，对备选模型进行超参数调优

#基于optimal特征子集，进行多模型集成
#why choose one model, when you can pick them all with voting classifier
#http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.VotingClassifier.html
#removed models w/o attribute 'predict_proba' required for vote classifier and models with a 1.0 correlation to another model
vote_est = [#Ensemble Methods: http://scikit-learn.org/stable/modules/ensemble.html('ada', ensemble.AdaBoostClassifier()),('rfc', ensemble.RandomForestClassifier()),('gbc', ensemble.GradientBoostingClassifier()),('xgb', XGBClassifier())#('bc', ensemble.BaggingClassifier()),#('etc',ensemble.ExtraTreesClassifier()),#('gbc', ensemble.GradientBoostingClassifier()),#('rfc', ensemble.RandomForestClassifier()),#Gaussian Processes: http://scikit-learn.org/stable/modules/gaussian_process.html#gaussian-process-classification-gpc#('gpc', gaussian_process.GaussianProcessClassifier()),#GLM: http://scikit-learn.org/stable/modules/linear_model.html#logistic-regression#('lr', linear_model.LogisticRegressionCV()),#Navies Bayes: http://scikit-learn.org/stable/modules/naive_bayes.html#('bnb', naive_bayes.BernoulliNB()),#('gnb', naive_bayes.GaussianNB()),#Nearest Neighbor: http://scikit-learn.org/stable/modules/neighbors.html#('knn', neighbors.KNeighborsClassifier()),#SVM: http://scikit-learn.org/stable/modules/svm.html# ('svc', svm.SVC(probability=True)),#xgboost: http://xgboost.readthedocs.io/en/latest/model.html#('xgb', XGBClassifier())]#WARNING: Running is very computational intensive and time expensive.
#Code is written for experimental/developmental purposes and not production ready!
#Hyperparameter Tune with GridSearchCV: http://scikit-learn.org/stable/modules/generated/sklearn.model_selection.GridSearchCV.html
grid_param = [[{#AdaBoostClassifier - http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.AdaBoostClassifier.html'n_estimators': grid_n_estimator, #default=50'learning_rate': grid_learn, #default=1#'algorithm': ['SAMME', 'SAMME.R'], #default=’SAMME.R'random_state': grid_seed}],[{#RandomForestClassifier - http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestClassifier.html#sklearn.ensemble.RandomForestClassifier'n_estimators': grid_n_estimator, #default=10'criterion': grid_criterion, #default=”gini”'max_depth': grid_max_depth, #default=None'oob_score': [True], #default=False -- 12/31/17 set to reduce runtime -- The best parameter for RandomForestClassifier is {'criterion': 'entropy', 'max_depth': 6, 'n_estimators': 100, 'oob_score': True, 'random_state': 0} with a runtime of 146.35 seconds.'random_state': grid_seed}],[{#GradientBoostingClassifier - http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingClassifier.html#sklearn.ensemble.GradientBoostingClassifier#'loss': ['deviance', 'exponential'], #default=’deviance’'learning_rate': [.05], #default=0.1 -- 12/31/17 set to reduce runtime -- The best parameter for GradientBoostingClassifier is {'learning_rate': 0.05, 'max_depth': 2, 'n_estimators': 300, 'random_state': 0} with a runtime of 264.45 seconds.'n_estimators': [300], #default=100 -- 12/31/17 set to reduce runtime -- The best parameter for GradientBoostingClassifier is {'learning_rate': 0.05, 'max_depth': 2, 'n_estimators': 300, 'random_state': 0} with a runtime of 264.45 seconds.#'criterion': ['friedman_mse', 'mse', 'mae'], #default=”friedman_mse”'max_depth': grid_max_depth, #default=3   'random_state': grid_seed}],[{#XGBClassifier - http://xgboost.readthedocs.io/en/latest/parameter.html'learning_rate': grid_learn, #default: .3'max_depth': [1,2,4,6,8,10], #default 2'n_estimators': grid_n_estimator, 'seed': grid_seed  }] ,'''[{#ExtraTreesClassifier - http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.ExtraTreesClassifier.html#sklearn.ensemble.ExtraTreesClassifier'n_estimators': grid_n_estimator, #default=10'criterion': grid_criterion, #default=”gini”'max_depth': grid_max_depth, #default=None'random_state': grid_seed}],[{#BaggingClassifier - http://scikit-learn.org/stable/modules/generated/sklearn.ensemble.BaggingClassifier.html#sklearn.ensemble.BaggingClassifier'n_estimators': grid_n_estimator, #default=10'max_samples': grid_ratio, #default=1.0'random_state': grid_seed}],[{    #GaussianProcessClassifier'max_iter_predict': grid_n_estimator, #default: 100'random_state': grid_seed}],[{#LogisticRegressionCV - http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegressionCV.html#sklearn.linear_model.LogisticRegressionCV'fit_intercept': grid_bool, #default: True#'penalty': ['l1','l2'],'solver': ['newton-cg', 'lbfgs', 'liblinear', 'sag', 'saga'], #default: lbfgs'random_state': grid_seed}],[{#BernoulliNB - http://scikit-learn.org/stable/modules/generated/sklearn.naive_bayes.BernoulliNB.html#sklearn.naive_bayes.BernoulliNB'alpha': grid_ratio, #default: 1.0}],#GaussianNB - [{}],[{#KNeighborsClassifier - http://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsClassifier.html#sklearn.neighbors.KNeighborsClassifier'n_neighbors': [1,2,3,4,5,6,7], #default: 5'weights': ['uniform', 'distance'], #default = ‘uniform’'algorithm': ['auto', 'ball_tree', 'kd_tree', 'brute']}],[{#SVC - http://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html#sklearn.svm.SVC#http://blog.hackerearth.com/simple-tutorial-svm-parameter-tuning-python-r#'kernel': ['linear', 'poly', 'rbf', 'sigmoid'],'C': [1,2,3,4,5], #default=1.0'gamma': grid_ratio, #edfault: auto'decision_function_shape': ['ovo', 'ovr'], #default:ovr'probability': [True],'random_state': grid_seed}],''']start_total = time.perf_counter() #https://docs.python.org/3/library/time.html#time.perf_counter
for clf, param in zip (vote_est, grid_param): #https://docs.python.org/3/library/functions.html#zip#print(clf[0],clf[1]) #vote_est is a list of tuples, index 0 is the name and index 1 is the algorithm#print(param)start = time.perf_counter()        #best_search = model_selection.GridSearchCV(estimator = clf[1], param_grid = param, cv = cv_split, scoring = SCORE_EVA)best_search = BayesSearchCV(clf[1], search_spaces=param, scoring=SCORE_EVA, cv=cv_split, n_iter=50,n_jobs=16)best_search.fit(X_best, y)run = time.perf_counter() - startbest_param = best_search.best_params_clf[1].set_params(**best_param) #对模型进行多次评估，以分析泛化能力scores = model_evaluate(X_best, y,clf[1],10,test_size=0.3)print('The best parameter for {} is {} with a runtime of {:.2f} seconds, scoreing is  {:.3f}, std: {:.3f}'.format(clf[1].__class__.__name__, best_param, run,np.mean(scores),np.std(scores)))run_total = time.perf_counter() - start_total
print('Total optimization time was {:.2f} minutes.'.format(run_total/60))
print('-'*10)

5.完成各模型超参数调优后，进行多模型集成，包括ensemble 或 stacking

#通过投票法 进行多模型集成
#Soft Vote or weighted probabilities w/Tuned Hyperparameters
vote = ensemble.VotingClassifier(estimators = vote_est , voting = 'soft')  #voting = 'hard'
vote_cv = model_selection.cross_validate(vote, X_best, y, cv  = cv_split,scoring = SCORE_EVA,return_train_score=True,n_jobs=16)
print("Soft Voting Training w/bin score mean: {:.2f}". format(vote_cv['train_score'].mean()*100)) 
print("Soft Voting Test w/bin score mean: {:.2f}". format(vote_cv['test_score'].mean()*100))
print("Soft Voting Test w/bin score 3*std: +/- {:.2f}". format(vote_cv['test_score'].std()*100*3))
print('-'*10)# stacking 法
meta_learner = LogisticRegression()
stacking_model = StackingClassifier(estimators=vote_est, final_estimator=meta_learner)
stacking_cv = model_selection.cross_validate(stacking_model, X_best, y, cv  = cv_split,scoring = SCORE_EVA,return_train_score=True,n_jobs=16)
print("Stacking Training w/bin score mean: {:.2f}". format(stacking_cv['train_score'].mean()*100)) 
print("Stacking Test w/bin score mean: {:.2f}". format(stacking_cv['test_score'].mean()*100))
print("Stacking Test w/bin score 3*std: +/- {:.2f}". format(stacking_cv['test_score'].std()*100*3))
print('-'*10)

参考： A Data Science Framework: To Achieve 99% Accuracy | Kaggle

本文来自互联网用户投稿，该文观点仅代表作者本人，不代表本站立场。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如若转载，请注明出处：http://www.mzph.cn/news/799298.shtml

如若内容造成侵权/违法违规/事实不符，请联系多彩编程网进行投诉反馈email:809451989@qq.com，一经查实，立即删除！

相关文章

（013）window的Idea运行程序 Amazon java.nio.file.AccessDeniedException

（013）window的Idea运行程序 Amazon java.nio.file.AccessDeniedException

解决方法一 在资源管理器中删除该目录， 在程序中使用代码，重新建立该目录： if (!FileUtil.exist(destinationPath)){FileUtil.mkdir(destinationPath); }解决方法二 JDK 的版本有问题，换个JDK。 解决方法三 网络不好&#xf…
阅读更多...
「39」打造专业流畅的直播特效转场……

「39」打造专业流畅的直播特效转场……

「39」工作室模式 打造专业流畅的直播特效转场体验 工作室模式是OBS软件里的一个特殊功能，用于后期直播过程中追求直播效果的用户，才会使用此功能。 该功能意在更加平滑，使用模板信息变化的过渡效果。主要用在赛事比分、活动抽奖、直播时需要经常更改的场景和内容，以及片…
阅读更多...
Spring之底层架构核心概念解析

Spring之底层架构核心概念解析

你好，我是柳岸花开。 在当今快速发展的软件开发领域，各种技术层出不穷，但有一项技术以其卓越的设计、灵活的配置和广泛的应用，始终屹立在开发者的工具箱中——这就是Spring框架。自2003年首次发布以来，Spring已经成为J…
阅读更多...
Android 11属性系统初始化流程

Android 11属性系统初始化流程

在init进程启动的第二阶段，调用PropertyInit 对属性系统进行初始化 int SecondStageMain(int argc, char** argv) {//省略PropertyInit();//省略 }PropertyInit函数在system\core\init\property_service.cpp 中实现 void PropertyInit() {//省略mkdir("/dev/…
阅读更多...
opencv x86(32位) windows下vs2019编译问题

opencv x86(32位) windows下vs2019编译问题

opencv x86(32位) windows下vs2019编译相关资料很多，都特别受用。例如 - https://blog.csdn.net/m0_59025104/article/details/134109081 - https://blog.csdn.net/sements/article/details/108410470 但是，自己编译时候仍遇到一些问题，例如&…
阅读更多...
智能网联汽车自动驾驶数据记录系统DSSAD数据元素

智能网联汽车自动驾驶数据记录系统DSSAD数据元素

目录 第一章 数据元素分级 第二章 数据元素分类 第三章 数据元素基本信息表 表1 车辆及自动驾驶数据记录系统基本信息 表2 车辆状态及动态信息 表3 自动驾驶系统运行信息 表4 行车环境信息 表5 驾驶员操作及状态信息 第一章 数据元素分级 自动驾驶数据记录系统记录的数…
阅读更多...
【GameFi】 链游 | Brilliantcrypto点火活动

【GameFi】 链游 | Brilliantcrypto点火活动

活动：https://app.galxe.com/quest/brilliantcrypto/GCt8wthq2J Brilliantcrypto点火活动正在Galxe上进行 🎉 活动时间：2024/04/06 12:00 ~ 2024/05/04 12:00 奖励总价值1200美元的MATIC 完成任务並在Brilliantcrypto Galxe Space上赚取积分。…
阅读更多...
【数据结构与算法】：快速排序和冒泡排序

【数据结构与算法】：快速排序和冒泡排序

一，快速排序 快速排序是一种比较复杂的排序算法，它总共有4种实现方式，分别是挖坑法，左右"指针"法，前后"指针"法，以及非递归的快速排序，并且这些算法中也会涉及多种优化措施…
阅读更多...
Docker是一个开源的应用容器引擎

Docker是一个开源的应用容器引擎

Docker是一个开源的应用容器引擎，它让开发者可以打包他们的应用以及依赖包到一个可移植的容器中，然后发布到任何Linux或Windows操作系统的机器上，也可以实现虚拟化。容器是完全使用沙箱机制，相互之间不会有任何接口。Docker技术的…
阅读更多...
zdpdjango_materialadmin使用Django开发一个Material风格的后台管理系统

zdpdjango_materialadmin使用Django开发一个Material风格的后台管理系统

启动项目 同样地，还是先看看代码： 将项目启动起来： 浏览器访问：http://localhost:8000/ 代码初步优化 首先是将admin_materal提到本地来： 移除掉第三方依赖： pip uninstall django-admin-materi…
阅读更多...
sqlserver2014安装与老版本卸载

sqlserver2014安装与老版本卸载

本篇博客只提供安装包 安装与卸载难度较低，不做赘述 并不是说一定要卸载才能安装，灵活掌握，如果实际场景需要清理老版本sqlserver那么选择卸载 一、下载 下载地址 sqlserver2014官方版下载丨最新版下载丨绿色版下载丨APP下载-123云盘 二、卸载…
阅读更多...
移动平台相关（安卓）

移动平台相关（安卓）

目录 安卓开发 Unity打包安卓 ​编辑​编辑 BuildSettings PlayerSettings OtherSettings 身份证明 配置 脚本编译 优化 PublishingSettings 调试 ReMote Android Logcat AndroidStudio的调试 Java语法 ​编辑​编辑​编辑 变量 运算符 ​编辑​编辑​编辑​…
阅读更多...
面向低碳经济运行目标的多微网能量互联优化调度matlab程序

面向低碳经济运行目标的多微网能量互联优化调度matlab程序

微❤关注“电气仔推送”获得资料（专享优惠） 运用平台 matlabgurobi 程序简介 该程序为多微网协同优化调度模型，系统在保障综合效益的基础上，调度时优先协调微网与微网之间的能量流动，将与大电网的互联交互作为备用…
阅读更多...
gcc/g++：预编译阶段查看模块生成目标的层级依赖

gcc/g++：预编译阶段查看模块生成目标的层级依赖

预编译阶段查看模块生成目标的层级依赖首先需要找到需要包含头文件的位置，然后进行引入。 示例： 1）用户头文件 /*brief design and implements of demo-for-precompile.author wenxuanpeiemail 15873152445163.com(query for any question …
阅读更多...
百度松果菁英班——机器学习实践四：文本词频分析

百度松果菁英班——机器学习实践四：文本词频分析

飞桨AI Studio星河社区-人工智能学习与实训社区 🥪jieba分词词频统计 import jieba # jieba中文分词库 ​ with open(test.txt, r, encodingUTF-8) as novelFile:novel novelFile.read() # print(novel) stopwords [line.strip() for line in open(stop.txt, r,…
阅读更多...
初识C++ · 类和对象(上)

初识C++ · 类和对象(上)

目录 1.面向过程和面向对象初步认识 2.类的引入 3.类的定义 4.类的访问限定符及封装 4.1 访问限定符 4.2 封装 5.类的作用域 6.类的实例化 7.类的对象大小的计算 8.类成员函数的this指针 1.面向过程和面向对象初步认识 C语言是一门面向过程的语言，注重的…
阅读更多...
vue+springboot多角色登录

vue+springboot多角色登录

①前端编写 将Homeview修改为manager Manager&#xff1a; <template><div><el-container><!-- 侧边栏 --><el-aside :width"asideWidth" style"min-height: 100vh; background-color: #001529"><div style"h…
阅读更多...
百度文库验证码识别

百度文库验证码识别

一、前言 百度出了如图所示的验证码&#xff0c;需要拖动滑块&#xff0c;与如图所示的曲线轨迹进行重合。经过不断研究&#xff0c;终于解决了这个问题。我把识别代码分享给大家。 下面是使用selenium进行验证的&#xff0c;这样可以看到轨迹滑动的过程&#xff0c;如果需要…
阅读更多...
亚马逊店铺引流：海外云手机的利用方法

亚马逊店铺引流：海外云手机的利用方法

在电商业务蓬勃发展的当下&#xff0c;亚马逊已经成为全球最大的电商平台之一&#xff0c;拥有庞大的用户群和交易量。在激烈的市场竞争中&#xff0c;如何有效地吸引流量成为亚马逊店铺经营者所关注的重点。海外云手机作为一项新兴技术工具&#xff0c;为亚马逊店铺的流量引导…
阅读更多...
页面转word的那些事

页面转word的那些事

背景 有些时候需要将页面内容或者是页面的数据通过word进行下载&#xff0c;以方便客户进行二次编辑&#xff0c;而不是直接导出图片或者是pdf。 想在页面端点击下载成word&#xff0c;那必然需要服务端来进行读写文件&#xff0c;无论是你后端编辑好的内容流&#xff0c;还是…
阅读更多...
最新文章

Copyright @ 2022~2023