課程完整代碼：https://github.com/fengdu78/WZU-machine-learning-course

代碼修改并注釋：黃海廣，haiguang2000@wzu.edu.cn

import?warnings warnings.filterwarnings("ignore") import?pandas?as?pd from?sklearn.model_selection?import?train_test_split

生成數(shù)據(jù)

生成12000行的數(shù)據(jù)，訓(xùn)練集和測(cè)試集按照3:1劃分

from?sklearn.datasets?import?make_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))

模型對(duì)比

對(duì)比六大模型，都使用默認(rèn)參數(shù)，因?yàn)閿?shù)據(jù)是

from?sklearn.linear_model?import?LogisticRegression from?sklearn.ensemble?import?RandomForestClassifier from?sklearn.ensemble?import?AdaBoostClassifier from?sklearn.ensemble?import?GradientBoostingClassifier from?xgboost?import?XGBClassifier from?lightgbm?import?LGBMClassifier from?sklearn.model_selection?import?cross_val_score import?timeclf1?=?LogisticRegression() clf2?=?RandomForestClassifier() clf3?=?AdaBoostClassifier() clf4?=?GradientBoostingClassifier() clf5?=?XGBClassifier() clf6?=?LGBMClassifier()for?clf,?label?in?zip([clf1,?clf2,?clf3,?clf4,?clf5,?clf6],?['Logistic?Regression',?'Random?Forest',?'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),耗時(shí)%0.2f秒。模型名稱[%s]"?%(scores.mean(),?scores.std(),?running_time,?label))Accuracy: 0.47488889 (+/- 0.00),耗時(shí)0.04秒。模型名稱[Logistic Regression] Accuracy: 0.88966667 (+/- 0.01),耗時(shí)16.34秒。模型名稱[Random Forest] Accuracy: 0.88311111 (+/- 0.00),耗時(shí)3.39秒。模型名稱[AdaBoost] Accuracy: 0.91388889 (+/- 0.01),耗時(shí)13.14秒。模型名稱[GBDT] Accuracy: 0.92977778 (+/- 0.00),耗時(shí)3.60秒。模型名稱[XGBoost] Accuracy: 0.93188889 (+/- 0.01),耗時(shí)0.58秒。模型名稱[LightGBM]

對(duì)比了六大模型，可以看出，邏輯回歸速度最快，但準(zhǔn)確率最低。而LightGBM，速度快，而且準(zhǔn)確率最高，所以，現(xiàn)在處理結(jié)構(gòu)化數(shù)據(jù)的時(shí)候，大部分都是用LightGBM算法。

XGBoost的使用

1.原生XGBoost的使用

import?xgboost?as?xgb #記錄程序運(yùn)行時(shí)間 import?timestart_time?=?time.time()#xgb矩陣賦值 xgb_train?=?xgb.DMatrix(X_train,?y_train) xgb_test?=?xgb.DMatrix(X_test,?label=y_test) ##參數(shù) params?=?{'booster':?'gbtree', #?????'silent':?1,??#設(shè)置成1則沒有運(yùn)行信息輸出，最好是設(shè)置為0.#'nthread':7,#?cpu?線程數(shù)?默認(rèn)最大'eta':?0.007,??#?如同學(xué)習(xí)率'min_child_weight':?3,#?這個(gè)參數(shù)默認(rèn)是?1，是每個(gè)葉子里面?h?的和至少是多少，對(duì)正負(fù)樣本不均衡時(shí)的?0-1?分類而言#，假設(shè) h 在?0.01 附近，min_child_weight 為 1 意味著葉子節(jié)點(diǎn)中最少需要包含 100?個(gè)樣本。#這個(gè)參數(shù)非常影響結(jié)果，控制葉子節(jié)點(diǎn)中二階導(dǎo)的和的最小值，該參數(shù)值越小，越容易 overfitting。'max_depth':?6,??#?構(gòu)建樹的深度，越大越容易過擬合'gamma':?0.1,??#?樹的葉子節(jié)點(diǎn)上作進(jìn)一步分區(qū)所需的最小損失減少,越大越保守，一般0.1、0.2這樣子。'subsample':?0.7,??#?隨機(jī)采樣訓(xùn)練樣本'colsample_bytree':?0.7,??#?生成樹時(shí)進(jìn)行的列采樣?'lambda':?2,??#?控制模型復(fù)雜度的權(quán)重值的L2正則化項(xiàng)參數(shù)，參數(shù)越大，模型越不容易過擬合。#'alpha':0,?#?L1?正則項(xiàng)參數(shù)#'scale_pos_weight':1, #如果取值大于0的話，在類別樣本不平衡的情況下有助于快速收斂。#'objective':?'multi:softmax',?#多分類的問題#'num_class':10,?#?類別數(shù)，多分類與?multisoftmax?并用'seed':?1000,??#隨機(jī)種子#'eval_metric':?'auc' } plst?=?list(params.items()) num_rounds?=?500??#?迭代次數(shù) watchlist?=?[(xgb_train,?'train'),?(xgb_test,?'val')]#訓(xùn)練模型并保存 #?early_stopping_rounds?當(dāng)設(shè)置的迭代次數(shù)較大時(shí)，early_stopping_rounds?可在一定的迭代次數(shù)內(nèi)準(zhǔn)確率沒有提升就停止訓(xùn)練 model?=?xgb.train(plst,xgb_train,num_rounds,watchlist,early_stopping_rounds=100, ) #model.save_model('./model/xgb.model')?#?用于存儲(chǔ)訓(xùn)練出的模型 print("best?best_ntree_limit",?model.best_ntree_limit) y_pred?=?model.predict(xgb_test,?ntree_limit=model.best_ntree_limit) print('error=%f'?%(sum(1for?i?in?range(len(y_pred))?if?int(y_pred[i]?>?0.5)?!=?y_test[i])?/float(len(y_pred)))) #?輸出運(yùn)行時(shí)長(zhǎng) cost_time?=?time.time()?-?start_time print("xgboost?success!",?'\n',?"cost?time:",?cost_time,?"(s)......")[0] train-rmse:1.11000 val-rmse:1.10422 [1] train-rmse:1.10734 val-rmse:1.10182 [2] train-rmse:1.10465 val-rmse:1.09932 [3] train-rmse:1.10207 val-rmse:1.09694…… [497] train-rmse:0.62135 val-rmse:0.68680 [498] train-rmse:0.62096 val-rmse:0.68650 [499] train-rmse:0.62056 val-rmse:0.68624 best best_ntree_limit 500 error=0.826667 xgboost success! cost time: 3.5742645263671875 (s)......

2.使用scikit-learn接口

會(huì)改變的函數(shù)名是：

eta -> learning_rate

lambda -> reg_lambda

alpha -> reg_alpha

from?sklearn.model_selection?import?train_test_split from?sklearn?import?metricsfrom?xgboost?import?XGBClassifierclf?=?XGBClassifier(#???? silent=0, ?#設(shè)置成1則沒有運(yùn)行信息輸出，最好是設(shè)置為0.是否在運(yùn)行升級(jí)時(shí)打印消息。#nthread=4,#?cpu?線程數(shù)?默認(rèn)最大learning_rate=0.3,??#?如同學(xué)習(xí)率min_child_weight=1,#?這個(gè)參數(shù)默認(rèn)是?1，是每個(gè)葉子里面?h?的和至少是多少，對(duì)正負(fù)樣本不均衡時(shí)的?0-1?分類而言#，假設(shè) h 在?0.01 附近，min_child_weight 為 1 意味著葉子節(jié)點(diǎn)中最少需要包含 100?個(gè)樣本。#這個(gè)參數(shù)非常影響結(jié)果，控制葉子節(jié)點(diǎn)中二階導(dǎo)的和的最小值，該參數(shù)值越小，越容易 overfitting。max_depth=6,??#?構(gòu)建樹的深度，越大越容易過擬合gamma=0,??#?樹的葉子節(jié)點(diǎn)上作進(jìn)一步分區(qū)所需的最小損失減少,越大越保守，一般0.1、0.2這樣子。subsample=1,??#?隨機(jī)采樣訓(xùn)練樣本?訓(xùn)練實(shí)例的子采樣比max_delta_step=0,??#最大增量步長(zhǎng)，我們?cè)试S每個(gè)樹的權(quán)重估計(jì)。colsample_bytree=1,??#?生成樹時(shí)進(jìn)行的列采樣?reg_lambda=1,??#?控制模型復(fù)雜度的權(quán)重值的L2正則化項(xiàng)參數(shù)，參數(shù)越大，模型越不容易過擬合。#reg_alpha=0,?#?L1?正則項(xiàng)參數(shù)#scale_pos_weight=1, #如果取值大于0的話，在類別樣本不平衡的情況下有助于快速收斂。平衡正負(fù)權(quán)重#objective=?'multi:softmax',?#多分類的問題?指定學(xué)習(xí)任務(wù)和相應(yīng)的學(xué)習(xí)目標(biāo)#num_class=10,?#?類別數(shù)，多分類與?multisoftmax?并用n_estimators=100,??#樹的個(gè)數(shù)seed=1000??#隨機(jī)種子#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))[20:16:02] WARNING: C:/Users/Administrator/workspace/xgboost-win64_release_1.3.0/src/learner.cc:1061: Starting in XGBoost 1.3.0, the default evaluation metric used with the objective 'binary:logistic' was changed from 'error' to 'logloss'. Explicitly set eval_metric if you'd like to restore the old behavior. Accuracy : 0.936

LIghtGBM的使用

1.原生接口

import?lightgbm?as?lgb from?sklearn.metrics?import?mean_squared_error #?加載你的數(shù)據(jù) #?print('Load?data...') #?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#?創(chuàng)建成lgb特征的數(shù)據(jù)集格式 lgb_train?=?lgb.Dataset(X_train,?y_train)??#?將數(shù)據(jù)保存到LightGBM二進(jìn)制文件將使加載更快 lgb_eval?=?lgb.Dataset(X_test,?y_test,?reference=lgb_train)??#?創(chuàng)建驗(yàn)證數(shù)據(jù)#?將參數(shù)寫成字典下形式 params?=?{'task':?'train','boosting_type':?'gbdt',??#?設(shè)置提升類型'objective':?'regression',??#?目標(biāo)函數(shù)'metric':?{'l2',?'auc'},??#?評(píng)估函數(shù)'num_leaves':?31,??#?葉子節(jié)點(diǎn)數(shù)'learning_rate':?0.05,??#?學(xué)習(xí)速率'feature_fraction':?0.9,??#?建樹的特征選擇比例'bagging_fraction':?0.8,??#?建樹的樣本采樣比例'bagging_freq':?5,??#?k?意味著每?k?次迭代執(zhí)行bagging'verbose':?1??#?<0?顯示致命的,?=0?顯示錯(cuò)誤?(警告),?>0?顯示信息 }print('Start?training...') #?訓(xùn)練?cv?and?train gbm?=?lgb.train(params,lgb_train,num_boost_round=500,valid_sets=lgb_eval,early_stopping_rounds=5)??#?訓(xùn)練數(shù)據(jù)需要參數(shù)列表和數(shù)據(jù)集print('Save?model...')gbm.save_model('model.txt')??#?訓(xùn)練后保存模型到文件print('Start?predicting...') #?預(yù)測(cè)數(shù)據(jù)集 y_pred?=?gbm.predict(X_test,?num_iteration=gbm.best_iteration)??#如果在訓(xùn)練期間啟用了早期停止，可以通過best_iteration方式從最佳迭代中獲得預(yù)測(cè) #?評(píng)估模型 print('error=%f'?%(sum(1for?i?in?range(len(y_pred))?if?int(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: 0.814399 valid_0's l2: 0.965563 Training until validation scores don't improve for 5 rounds [2] valid_0's auc: 0.84729 valid_0's l2: 0.934647 …… [193] valid_0's auc: 0.982685 valid_0's l2: 0.320043 Early stopping, best iteration is: [188] valid_0's auc: 0.982794 valid_0's l2: 0.319746 Save model... Start predicting... error=0.664000

2.scikit-learn接口

from?sklearn?import?metrics from?lightgbm?import?LGBMClassifierclf?=?LGBMClassifier(boosting_type='gbdt',??#?提升樹的類型?gbdt,dart,goss,rfnum_leaves=31,??#樹的最大葉子數(shù)，對(duì)比xgboost一般為2^(max_depth)max_depth=-1,??#最大樹的深度learning_rate=0.1,??#學(xué)習(xí)率n_estimators=100,??#?擬合的樹的棵樹，相當(dāng)于訓(xùn)練輪數(shù)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,??#?分支結(jié)點(diǎn)的最小權(quán)重min_child_samples=20,subsample=1.0,??#?訓(xùn)練樣本采樣率?行subsample_freq=0,??#?子樣本頻率colsample_bytree=1.0,??#?訓(xùn)練特征采樣率?列reg_alpha=0.0,??#?L1正則化系數(shù)reg_lambda=0.0,??#?L2正則化系數(shù)random_state=None,n_jobs=-1,silent=True, ) clf.fit(X_train,?y_train,?eval_metric='auc') #設(shè)置驗(yàn)證集合?verbose=False不打印過程 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

參考

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

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

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

往期精彩回顧適合初學(xué)者入門人工智能的路線及資料下載機(jī)器學(xué)習(xí)及深度學(xué)習(xí)筆記等資料打印機(jī)器學(xué)習(xí)在線手冊(cè)深度學(xué)習(xí)筆記專輯《統(tǒng)計(jì)學(xué)習(xí)方法》的代碼復(fù)現(xiàn)專輯 AI基礎(chǔ)下載黃海廣老師《機(jī)器學(xué)習(xí)課程》視頻課黃海廣老師《機(jī)器學(xué)習(xí)課程》711頁完整版課件

本站qq群554839127，加入微信群請(qǐng)掃碼：

總結(jié)

以上是生活随笔為你收集整理的【机器学习】集成学习代码练习的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。