【机器学习】集成学习代码练习（随机森林、GBDT、XGBoost、LightGBM等）

本文是中国大学慕课《机器学习》的“集成学习”章节的课后代码。

课程地址：

https://www.icourse163.org/course/WZU-1464096179

(资料图)

课程完整代码：

https://github.com/fengdu78/WZU-machine-learning-course

代码修改并注释：黄海广，haiguang2000@wzu.edu.cn

importwarningswarnings.filterwarnings("ignore")importpandasaspdfromsklearn.model_selectionimporttrain_test_split

生成12000行的数据，训练集和测试集按照3:1划分

fromsklearn.datasetsimportmake_hastie_10_2data,target=make_hastie_10_2()

X_train,X_test,y_train,y_test=train_test_split(data,target,random_state=123)X_train.shape,X_test.shape

((9000, 10), (3000, 10))

对比六大模型，都使用默认参数

fromsklearn.linear_modelimportLogisticRegressionfromsklearn.ensembleimportRandomForestClassifierfromsklearn.ensembleimportAdaBoostClassifierfromsklearn.ensembleimportGradientBoostingClassifierfromxgboostimportXGBClassifierfromlightgbmimportLGBMClassifierfromsklearn.model_selectionimportcross_val_scoreimporttimeclf1=LogisticRegression()clf2=RandomForestClassifier()clf3=AdaBoostClassifier()clf4=GradientBoostingClassifier()clf5=XGBClassifier()clf6=LGBMClassifier()forclf,labelinzip([clf1,clf2,clf3,clf4,clf5,clf6],["LogisticRegression","RandomForest","AdaBoost","GBDT","XGBoost","LightGBM"]):start=time.time()scores=cross_val_score(clf,X_train,y_train,scoring="accuracy",cv=5)end=time.time()running_time=end-startprint("Accuracy:%0.8f (+/-%0.2f),耗时%0.2f秒。模型名称[%s]"%(scores.mean(),scores.std(),running_time,label))

Accuracy: 0.47488889 (+/- 0.00),耗时0.04秒。模型名称[Logistic Regression]Accuracy: 0.88966667 (+/- 0.01),耗时16.34秒。模型名称[Random Forest]Accuracy: 0.88311111 (+/- 0.00),耗时3.39秒。模型名称[AdaBoost]Accuracy: 0.91388889 (+/- 0.01),耗时13.14秒。模型名称[GBDT]Accuracy: 0.92977778 (+/- 0.00),耗时3.60秒。模型名称[XGBoost]Accuracy: 0.93188889 (+/- 0.01),耗时0.58秒。模型名称[LightGBM]

对比了六大模型，可以看出，逻辑回归速度最快，但准确率最低。而LightGBM，速度快，而且准确率最高，所以，现在处理结构化数据的时候，大部分都是用LightGBM算法。

importxgboostasxgb#记录程序运行时间importtimestart_time=time.time()#xgb矩阵赋值xgb_train=xgb.DMatrix(X_train,y_train)xgb_test=xgb.DMatrix(X_test,label=y_test)##参数params={"booster":"gbtree",#"silent":1,#设置成1则没有运行信息输出，最好是设置为0.#"nthread":7,#cpu线程数默认最大"eta":0.007,#如同学习率"min_child_weight":3,#这个参数默认是1，是每个叶子里面h的和至少是多少，对正负样本不均衡时的0-1分类而言#，假设 h 在0.01 附近，min_child_weight 为 1 意味着叶子节点中最少需要包含 100个样本。#这个参数非常影响结果，控制叶子节点中二阶导的和的最小值，该参数值越小，越容易 overfitting。"max_depth":6,#构建树的深度，越大越容易过拟合"gamma":0.1,#树的叶子节点上作进一步分区所需的最小损失减少,越大越保守，一般0.1、0.2这样子。"subsample":0.7,#随机采样训练样本"colsample_bytree":0.7,#生成树时进行的列采样"lambda":2,#控制模型复杂度的权重值的L2正则化项参数，参数越大，模型越不容易过拟合。#"alpha":0,#L1正则项参数#"scale_pos_weight":1, #如果取值大于0的话，在类别样本不平衡的情况下有助于快速收敛。#"objective":"multi:softmax",#多分类的问题#"num_class":10,#类别数，多分类与multisoftmax并用"seed":1000,#随机种子#"eval_metric":"auc"}plst=list(params.items())num_rounds=500#迭代次数watchlist=[(xgb_train,"train"),(xgb_test,"val")]

#训练模型并保存#early_stopping_rounds当设置的迭代次数较大时，early_stopping_rounds可在一定的迭代次数内准确率没有提升就停止训练model=xgb.train(plst,xgb_train,num_rounds,watchlist,early_stopping_rounds=100,)#model.save_model("./model/xgb.model")#用于存储训练出的模型print("bestbest_ntree_limit",model.best_ntree_limit)y_pred=model.predict(xgb_test,ntree_limit=model.best_ntree_limit)print("error=%f"%(sum(1foriinrange(len(y_pred))ifint(y_pred[i]>0.5)!=y_test[i])/float(len(y_pred))))#输出运行时长cost_time=time.time()-start_timeprint("xgboostsuccess!","\n","costtime:",cost_time,"(s)......")

[0]train-rmse:1.11000val-rmse:1.10422[1]train-rmse:1.10734val-rmse:1.10182[2]train-rmse:1.10465val-rmse:1.09932[3]train-rmse:1.10207val-rmse:1.09694

……

[497]train-rmse:0.62135val-rmse:0.68680[498]train-rmse:0.62096val-rmse:0.68650[499]train-rmse:0.62056val-rmse:0.68624best best_ntree_limit 500error=0.826667xgboost success!  cost time: 3.5742645263671875 (s)......

会改变的函数名是：

eta -> learning_rate

lambda -> reg_lambda

alpha -> reg_alpha

fromsklearn.model_selectionimporttrain_test_splitfromsklearnimportmetricsfromxgboostimportXGBClassifierclf=XGBClassifier(# silent=0, #设置成1则没有运行信息输出，最好是设置为0.是否在运行升级时打印消息。#nthread=4,#cpu线程数默认最大learning_rate=0.3,#如同学习率min_child_weight=1,#这个参数默认是1，是每个叶子里面h的和至少是多少，对正负样本不均衡时的0-1分类而言#，假设 h 在0.01 附近，min_child_weight 为 1 意味着叶子节点中最少需要包含 100个样本。#这个参数非常影响结果，控制叶子节点中二阶导的和的最小值，该参数值越小，越容易 overfitting。max_depth=6,#构建树的深度，越大越容易过拟合gamma=0,#树的叶子节点上作进一步分区所需的最小损失减少,越大越保守，一般0.1、0.2这样子。subsample=1,#随机采样训练样本训练实例的子采样比max_delta_step=0,#最大增量步长，我们允许每个树的权重估计。colsample_bytree=1,#生成树时进行的列采样reg_lambda=1,#控制模型复杂度的权重值的L2正则化项参数，参数越大，模型越不容易过拟合。#reg_alpha=0,#L1正则项参数#scale_pos_weight=1, #如果取值大于0的话，在类别样本不平衡的情况下有助于快速收敛。平衡正负权重#objective="multi:softmax",#多分类的问题指定学习任务和相应的学习目标#num_class=10,#类别数，多分类与multisoftmax并用n_estimators=100,#树的个数seed=1000#随机种子#eval_metric="auc")clf.fit(X_train,y_train)y_true,y_pred=y_test,clf.predict(X_test)print("Accuracy:%.4g"%metrics.accuracy_score(y_true,y_pred))

Accuracy : 0.936

importlightgbmaslgbfromsklearn.metricsimportmean_squared_error#加载你的数据#print("Loaddata...")#df_train=pd.read_csv("../regression/regression.train",header=None,sep="\t")#df_test=pd.read_csv("../regression/regression.test",header=None,sep="\t")##y_train=df_train[0].values#y_test=df_test[0].values#X_train=df_train.drop(0,axis=1).values#X_test=df_test.drop(0,axis=1).values#创建成lgb特征的数据集格式lgb_train=lgb.Dataset(X_train,y_train)#将数据保存到LightGBM二进制文件将使加载更快lgb_eval=lgb.Dataset(X_test,y_test,reference=lgb_train)#创建验证数据#将参数写成字典下形式params={"task":"train","boosting_type":"gbdt",#设置提升类型"objective":"regression",#目标函数"metric":{"l2","auc"},#评估函数"num_leaves":31,#叶子节点数"learning_rate":0.05,#学习速率"feature_fraction":0.9,#建树的特征选择比例"bagging_fraction":0.8,#建树的样本采样比例"bagging_freq":5,#k意味着每k次迭代执行bagging"verbose":1#<0显示致命的,=0显示错误(警告),>0显示信息}print("Starttraining...")#训练cvandtraingbm=lgb.train(params,lgb_train,num_boost_round=500,valid_sets=lgb_eval,early_stopping_rounds=5)#训练数据需要参数列表和数据集print("Savemodel...")gbm.save_model("model.txt")#训练后保存模型到文件print("Startpredicting...")#预测数据集y_pred=gbm.predict(X_test,num_iteration=gbm.best_iteration)#如果在训练期间启用了早期停止，可以通过best_iteration方式从最佳迭代中获得预测#评估模型print("error=%f"%(sum(1foriinrange(len(y_pred))ifint(y_pred[i]>0.5)!=y_test[i])/float(len(y_pred))))

Start training...[LightGBM] [Warning] Auto-choosing col-wise multi-threading, the overhead of testing was 0.000448 seconds.You can set `force_col_wise=true` to remove the overhead.[LightGBM] [Info] Total Bins 2550[LightGBM] [Info] Number of data points in the train set: 9000, number of used features: 10[LightGBM] [Info] Start training from score 0.012000[1]valid_0"s auc: 0.814399valid_0"s l2: 0.965563Training until validation scores don"t improve for 5 rounds[2]valid_0"s auc: 0.84729valid_0"s l2: 0.934647[3]valid_0"s auc: 0.872805valid_0"s l2: 0.905265[4]valid_0"s auc: 0.884117valid_0"s l2: 0.877875[5]valid_0"s auc: 0.895115valid_0"s l2: 0.852189

……

[191]valid_0"s auc: 0.982783valid_0"s l2: 0.319851[192]valid_0"s auc: 0.982751valid_0"s l2: 0.319971[193]valid_0"s auc: 0.982685valid_0"s l2: 0.320043Early stopping, best iteration is:[188]valid_0"s auc: 0.982794valid_0"s l2: 0.319746Save model...Start predicting...error=0.664000

fromsklearnimportmetricsfromlightgbmimportLGBMClassifierclf=LGBMClassifier(boosting_type="gbdt",#提升树的类型gbdt,dart,goss,rfnum_leaves=31,#树的最大叶子数，对比xgboost一般为2^(max_depth)max_depth=-1,#最大树的深度learning_rate=0.1,#学习率n_estimators=100,#拟合的树的棵树，相当于训练轮数subsample_for_bin=200000,objective=None,class_weight=None,min_split_gain=0.0,#最小分割增益min_child_weight=0.001,#分支结点的最小权重min_child_samples=20,subsample=1.0,#训练样本采样率行subsample_freq=0,#子样本频率colsample_bytree=1.0,#训练特征采样率列reg_alpha=0.0,#L1正则化系数reg_lambda=0.0,#L2正则化系数random_state=None,n_jobs=-1,silent=True,)clf.fit(X_train,y_train,eval_metric="auc")#设置验证集合verbose=False不打印过程clf.fit(X_train,y_train)y_true,y_pred=y_test,clf.predict(X_test)print("Accuracy:%.4g"%metrics.accuracy_score(y_true,y_pred))

Accuracy : 0.927

1.https://xgboost.readthedocs.io/

2.https://lightgbm.readthedocs.io/

3.https://blog.csdn.net/q383700092/article/details/53763328?locationNum=9&fps=1

往期精彩回顾适合初学者入门人工智能的路线及资料下载(图文+视频)机器学习入门系列下载机器学习及深度学习笔记等资料打印《统计学习方法》的代码复现专辑机器学习交流qq群955171419，加入微信群请扫码

关键词：