python中的多随机回归模型

Question

在 X_train 集和 Y_train 标签上构建多个随机森林回归器，其中 max_depth 参数值从 3 变为 5，并将 n_estimators 设置为 50、100、200 值之一。在测试数据集上评估每个模型的准确性。提示：使用 for 循环以最高精度打印模型的 max_depth 和 n_estimators 值。注意：以元组（a，b）的形式打印参数值。a 指的是 max_depth 值，b 指的是 n_estimators到目前为止，这是我尝试过的：boston= datasets.load_boston()X_train, X_test, Y_train, Y_test = train_test_split(boston.data, boston.target, random_state=30)for m in range(3,6) :    rf_reg = RandomForestRegressor(n_estimators =100, max_depth=m)    rf_reg = rf_reg.fit(X_train, Y_train)     print(rf_reg.score(X_test,Y_test))这给了我 3 个模型的准确度分数，但我无法单独获取最高精度的参数。我可以使用rf_reg.get_params()，但它给了我所有的参数。我只想要max_depth和n_estimators最高分的一个

Answer 1

max_reg = None  #<--- add this to represent the regressor with maximum score

max_score = 0   #<--- add this to represent maximum score

t=() # <--- add this to tuple declaration

c_estimators = 100 

for m in range(3,6) :

    rf_reg = RandomForestRegressor(n_estimators =c_estimators , max_depth=m)

    rf_reg = rf_reg.fit(X_train, Y_train) 

    rf_reg_score = rf_reg.score(X_test,Y_test)

    t = (m,c_estimators,rf_reg.score) # tuple assignment

    rf_reg_score = t[2]

    print (t)

    if rf_reg_score > max_score :

        max_score = rf_reg_score

        max_reg = rf_reg

        t = (m,c_estimators) # tuple assignment

print (t)

Answer 2

您将获得 (5, 100) 组合的 max_score。根据问题，他们要求执行总共 9 种组合。3×3。

Answer 3

如何以元组 (a, b) 的形式打印参数值。a 指的是 max_depth 值，b 指的是 n_estimators

Answer 4

import sklearn.datasets as datasets

import sklearn.model_selection as model_selection

import numpy as np

from sklearn.model_selection import train_test_split

from sklearn.ensemble import RandomForestRegressor

np.random.seed(100)

boston=datasets.load_boston()

X_train,X_test,Y_train,Y_test=train_test_split(boston.data,boston.target,random_state=30)

print(X_train.shape)

print(X_test.shape)

rf_reg = RandomForestRegressor()

rf_reg = rf_reg.fit(X_train, Y_train)

print(rf_reg.score(X_train,Y_train))

print(rf_reg.score(X_test,Y_test))

print(rf_reg.predict(X_test[0:2]))

li=[]

nestimators=100

for maxdepth in range(3,6) :

    rf_reg1 = RandomForestRegressor(max_depth=maxdepth,n_estimators=nestimators)

    rf_reg1 = rf_reg1.fit(X_train, Y_train)  

    li.append(rf_reg1.score(X_test,Y_test))

maxValue=max(li)

maxIndex=li.index(maxValue)

a=(maxIndex+3,nestimators)

print(a)

#This code 100% works ,i tested and got exact output and cleared HandsOn