02-07 多元線性迴歸(波士頓房價預測)

[TOC] 更新、更全的《機器學習》的更新網站，更有python、go、數據結構與算法、爬蟲、人工智能教學等着你：http://www.javashuo.com/article/p-vozphyqp-cm.html

多元線性迴歸(波士頓房價預測)

1、導入模塊

import pandas as pd
import matplotlib.pyplot as plt
from matplotlib.font_manager import FontProperties
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
%matplotlib inline
font = FontProperties(fname='/Library/Fonts/Heiti.ttc')

2、獲取數據

df = pd.read_csv('housing-data.txt', sep='\s+', header=0)
X = df.iloc[:, :-1].values
y = df['MEDV'].values
# 將數據分紅訓練集（0.7）和測試集（0.3）
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3)

3、訓練模型

lr = LinearRegression()
# 訓練模型
lr.fit(X_train, y_train)
# 預測訓練集數據
y_train_predict = lr.predict(X_train)
# 預測測試集數據
y_test_predict = lr.predict(X_test)

4、可視化

# y_train_predict-y_train訓練數據偏差值
plt.scatter(y_train_predict, y_train_predict-y_train, c='r',
            marker='s', edgecolor='white', label='訓練數據')
# y_train_predict-y_train測試數據偏差值
plt.scatter(y_test_predict, y_test_predict-y_test, c='g',
            marker='o', edgecolor='white', label='測試數據')
plt.xlabel('預測值', fontproperties=font)
plt.ylabel('偏差值', fontproperties=font)
# 可視化y=0的一條直線即偏差爲0的直線
plt.hlines(y=0, xmin=-10, xmax=50, color='k')
plt.xlim(-10, 50)
plt.legend(prop=font)
plt.show()

_8_0.png?x-oss-process=style/watermark)

5、均方偏差測試

from sklearn.metrics import mean_squared_error

# 訓練集的均方偏差
train_mse = mean_squared_error(y_train,y_train_predict)
# 測試集的均方偏差
test_mse = mean_squared_error(y_test,y_test_predict)
print('訓練集的均方偏差:{}'.format(train_mse))
print('測試集的均方偏差:{}'.format(test_mse))

訓練集的均方偏差:23.049177061822277
測試集的均方偏差:19.901828312902534

訓練集的均方偏差是19.4，而測試集的均方偏差是28.4，能夠發現測試集的偏差更大了，也就是說訓練集過擬合了。