机器学习-多元线性回归

用python求解多元线性回归问题 

#1.导入工具包
import numpy as np
import pandas as pd

#2.读取数据
dt = pd.read_csv('breast_cancer_train.csv')
dt.head()

#3.拆分数据
#3.1 分离ID，分离label
X = dt.iloc[:,1:-1]
# X.shape
y = dt.iloc[:,-1]
# y.head()
# -------------------处理数据，归一化数据
# 1.导入工具包，处理函数
from sklearn.preprocessing import StandardScaler,MinMaxScaler
# 2.调用函数进行处理
# 2.2 初始化处理类
sd = StandardScaler()
# 2.2 训练类
sd.fit(X)
#2.3 转换数据
X = sd.transform(X)
# --------------------
#拆分数据
from sklearn.model_selection import train_test_split
# 3.2 拆分验证数据和训练数据
X_train,X_val,y_train,y_val=train_test_split(X,y,test_size=0.2,random_state=1,shuffle=True)
#print(X_train.shape,X_val.shape)

# 1.导入线行模型--根据不同场景选择不同的模型
from sklearn.linear_model import LogisticRegression
# 2.初始化线性模型(此处调参)
lr = LogisticRegression(penalty = 'l1',solver = 'liblinear',C = 1.5)
# 3.训练线下模型,在训练数据集上
lr.fit(X_train,y_train)
# 4.验证线性模型，在训练数据集上
print(lr.score(X_train,y_train))
# 5.在验证数据集上验证
print(lr.score(X_val,y_val))

# 1.读入测试数据
dt_tst = pd.read_csv('breast_cancer_test.csv')
# dt_tst
# 2.分离数据
y_id = dt_tst.iloc[:,0]
X_tst = dt_tst.iloc[:,1:]
# 3.归一化数据，使用训练数据的归一化模型进行归一化
X_tst = sd.transform(X_tst)
# 4.预测测试数据
y_pred = lr.predict(X_tst)
y_pred
# 5.保存数据
df=pd.DataFrame({'ID':y_id,r'label':y_pred})
df.to_csv('yours_path.csv')