机器学习sklearn（89）：算法实例（46）分类（25）XGBoost（三）梯度提升树（二）有放回随机抽样：重要参数subsample/迭代决策树：重要参数eta

1 有放回随机抽样：重要参数subsample

 

 

 

 

 

axisx = np.linspace(0,1,20)
rs = []
for i in axisx:
    reg = XGBR(n_estimators=180,subsample=i,random_state=420)
    rs.append(CVS(reg,Xtrain,Ytrain,cv=cv).mean())
print(axisx[rs.index(max(rs))],max(rs))
plt.figure(figsize=(20,5))
plt.plot(axisx,rs,c="green",label="XGB")
plt.legend()
plt.show()
#细化学习曲线
axisx = np.linspace(0.05,1,20)
rs = []
var = []
ge = []
for i in axisx:
    reg = XGBR(n_estimators=180,subsample=i,random_state=420)
    cvresult = CVS(reg,Xtrain,Ytrain,cv=cv)
    rs.append(cvresult.mean())
    var.append(cvresult.var())
    ge.append((1 - cvresult.mean())**2+cvresult.var())
print(axisx[rs.index(max(rs))],max(rs),var[rs.index(max(rs))])
print(axisx[var.index(min(var))],rs[var.index(min(var))],min(var))
print(axisx[ge.index(min(ge))],rs[ge.index(min(ge))],var[ge.index(min(ge))],min(ge))
rs = np.array(rs)
var = np.array(var)
plt.figure(figsize=(20,5))
plt.plot(axisx,rs,c="black",label="XGB")
plt.plot(axisx,rs+var,c="red",linestyle='-.')
plt.plot(axisx,rs-var,c="red",linestyle='-.')
plt.legend()
plt.show()
#继续细化学习曲线
axisx = np.linspace(0.75,1,25) #不要盲目找寻泛化误差可控部分的最低值，注意观察结果
#看看泛化误差的情况如何
reg = XGBR(n_estimators=180
           ,subsample=0.7708333333333334
           ,random_state=420).fit(Xtrain,Ytrain)
reg.score(Xtest,Ytest)
MSE(Ytest,reg.predict(Xtest))
#这样的结果说明了什么？

 2 迭代决策树：重要参数eta

 

 

 

 

 

#首先我们先来定义一个评分函数，这个评分函数能够帮助我们直接打印Xtrain上的交叉验证结果
def regassess(reg,Xtrain,Ytrain,cv,scoring = ["r2"],show=True):
    score = []
    for i in range(len(scoring)):
        if show:
            print("{}:{:.2f}".format(scoring[i]
                                     ,CVS(reg
                                         ,Xtrain,Ytrain
                                         ,cv=cv,scoring=scoring[i]).mean()))
        score.append(CVS(reg,Xtrain,Ytrain,cv=cv,scoring=scoring[i]).mean())
    return score
#运行一下函数来看看效果
regassess(reg,Xtrain,Ytrain,cv,scoring = ["r2","neg_mean_squared_error"])
#关闭打印功能试试看？
regassess(reg,Xtrain,Ytrain,cv,scoring = ["r2","neg_mean_squared_error"],show=False) #观察一下eta如何影响我们的模型：
from time import time
import datetime
for i in [0,0.2,0.5,1]:
    time0=time()
    reg = XGBR(n_estimators=180,random_state=420,learning_rate=i)
    print("learning_rate = {}".format(i))
    regassess(reg,Xtrain,Ytrain,cv,scoring = ["r2","neg_mean_squared_error"])
    print(datetime.datetime.fromtimestamp(time()-time0).strftime("%M:%S:%f"))
    print("	")

axisx = np.arange(0.05,1,0.05)
rs = []
te = []
for i in axisx:
    reg = XGBR(n_estimators=180,random_state=420,learning_rate=i)
    score = regassess(reg,Xtrain,Ytrain,cv,scoring = 
["r2","neg_mean_squared_error"],show=False)
    test = reg.fit(Xtrain,Ytrain).score(Xtest,Ytest)
    rs.append(score[0])
    te.append(test)
print(axisx[rs.index(max(rs))],max(rs))
plt.figure(figsize=(20,5))
plt.plot(axisx,te,c="gray",label="XGB")
plt.plot(axisx,rs,c="green",label="XGB")
plt.legend()
plt.show()