1.資料讀取

利用原生xgboost庫讀取libsvm資料

 import xgboost as xgb
 data = xgb.DMatrix(libsvm檔案)

使用sklearn讀取libsvm資料

 from sklearn.datasets import load_svmlight_file
 X_train,y_train = load_svmlight_file(libsvm檔案)

使用pandas讀取完資料後在轉化為標準形式

2.模型訓練過程

1.未調參基線模型

使用xgboost原生庫進行訓練

import xgboost as xgb
from sklearn.metrics import accuracy_score

dtrain = xgb.DMatrix(f_train,label = l_train)
dtest = xgb.DMatrix(f_test,label = l_test)
param = {'max_depth':2,'eta':1,'silent':0,'objective':'binary:logistic' }
num_round = 2
bst = xgb.train(param,dtrain,num_round)
train_preds = bst.predict(dtrain)
train_predictions = [round(value) for value in train_preds] #進行四捨五入的操作--變成0.1(算是設定閾值的符號函式)
train_accuracy = accuracy_score(l_train,train_predictions) #使用sklearn進行比較正確率
print ("Train Accuary: %.2f%%" % (train_accuracy * 100.0))

from xgboost import plot_importance #顯示特徵重要性
plot_importance(bst)#列印重要程度結果。
pyplot.show()

使用XGBClassifier進行訓練

# 未設定早停止， 未進行矩陣變換
from xgboost import XGBClassifier
from sklearn.datasets import load_svmlight_file #用於直接讀取svmlight檔案形式， 否則就需要使用xgboost.DMatrix(檔名)來讀取這種格式的檔案
from sklearn.metrics import accuracy_score
from matplotlib import pyplot


num_round = 100
bst1 =XGBClassifier(max_depth=2,learning_rate=1,n_estimators=num_round,#弱分類樹太少的話取不到更多的特徵重要性
          silent=True,objective='binary:logistic')
bst1.fit(f_train,l_train)

train_preds = bst1.predict(f_train)
train_accuracy = accuracy_score(l_train,train_preds)
print ("Train Accuary: %.2f%%" % (train_accuracy * 100.0))

preds = bst1.predict(f_test)
test_accuracy = accuracy_score(l_test,preds)
print("Test Accuracy: %.2f%%" % (test_accuracy * 100.0))

from xgboost import plot_importance #顯示特徵重要性
plot_importance(bst1)#列印重要程度結果。
pyplot.show()

2.兩種交叉驗證方式

使用cross_val_score進行交叉驗證

#利用model_selection進行交叉訓練
from xgboost import XGBClassifier
from sklearn.model_selection import StratifiedKFold
from sklearn.model_selection import cross_val_score
from sklearn.metrics import accuracy_score
from matplotlib import pyplot

param = {'max_depth':2,'objective':'binary:logistic' }
num_round = 100
bst2 =XGBClassifier(max_depth=2,learning_rate=0.1,silent=True,objective='binary:logistic')
bst2.fit(f_train,l_train)
kfold = StratifiedKFold(n_splits=10,random_state=7)
results = cross_val_score(bst2,f_train,l_train,cv=kfold)#對資料進行十折交叉驗證--9份訓練，一份測試
print(results)
print("CV Accuracy: %.2f%% (%.2f%%)" % (results.mean()*100,results.std()*100))

from xgboost import plot_importance #顯示特徵重要性
plot_importance(bst2)#列印重要程度結果。
pyplot.show()

使用GridSearchCV進行網格搜尋

#使用sklearn中提供的網格搜尋進行測試--找出最好引數，並作為預設訓練引數
from xgboost import XGBClassifier
from sklearn.model_selection import GridSearchCV
from sklearn.metrics import accuracy_score
from matplotlib import pyplot

params = {'max_depth':2,'eta':0.1,'objective':'binary:logistic' }
bst =XGBClassifier(max_depth=2,objective='binary:logistic')
param_test = {
 'n_estimators': range(1,51,1)
}
clf = GridSearchCV(estimator = bst,param_grid = param_test,scoring='accuracy',cv=5)# 5折交叉驗證
clf.fit(f_train,l_train) #預設使用最優的引數


preds = clf.predict(f_test)

test_accuracy = accuracy_score(l_test,preds)
print("Test Accuracy of gridsearchcv: %.2f%%" % (test_accuracy * 100.0))

clf.cv_results_,clf.best_params_,clf.best_score_
 

3.早停止調參–early_stopping_rounds（檢視的是損失是否變化）

#進行提早停止的單獨例項
import xgboost as xgb
from xgboost import XGBClassifier
from sklearn.metrics import accuracy_score
from matplotlib import pyplot

param = {'max_depth':2,'objective':'binary:logistic' }
num_round = 100
bst =XGBClassifier(max_depth=2,objective='binary:logistic')
eval_set =[(f_test,l_test)]
bst.fit(f_train,early_stopping_rounds=10,eval_metric="error",eval_set=eval_set,verbose=True) #early_stopping_rounds--當多少次的效果差不多時停止  eval_set--用於顯示損失率的資料 verbose--顯示錯誤率的變化過程

# make prediction
preds = bst.predict(f_test)

test_accuracy = accuracy_score(l_test,preds)
print("Test Accuracy: %.2f%%" % (test_accuracy * 100.0))

4.多資料觀察訓練損失

#多引數順
import xgboost as xgb
from xgboost import XGBClassifier
from sklearn.metrics import accuracy_score
from matplotlib import pyplot

num_round = 100
bst =XGBClassifier(max_depth=2,objective='binary:logistic')
eval_set = [(f_train,l_train),(f_test,eval_metric=["error","logloss"],verbose=True)

# make prediction
preds = bst.predict(f_test)
test_accuracy = accuracy_score(l_test,preds)
print("Test Accuracy: %.2f%%" % (test_accuracy * 100.0))

5.模型儲存與讀取

#模型儲存
bst.save_model('demo.model')

#模型讀取與預測
modelfile = 'demo.model'

# 1
bst = xgb.Booster({'nthread':8},model_file = modelfile)

# 2

f_test1 = xgb.DMatrix(f_test) #儘量使用xgboost的自己的資料矩陣
ypred1 = bst.predict(f_test1)
train_predictions = [round(value) for value in ypred1]
test_accuracy1 = accuracy_score(l_test,train_predictions)
print("Test Accuracy: %.2f%%" % (test_accuracy1 * 100.0))

以上就是本文的全部內容，希望對大家的學習有所幫助，也希望大家多多支援我們。