廣義線性模型

廣義線性模型有：最小二乘回歸、感知機(jī)、邏輯回歸、嶺回歸，貝葉斯回歸等，由 sklearn.linear_model 模塊導(dǎo)入。對(duì)于廣義線性模型而言，即通過(guò)擬合線性函數(shù)（下圖）去完成樣本分類(lèi)或回歸預(yù)測(cè)。

• 線性回歸

# -*- coding: utf-8 -*
from sklearn import linear_model
model = linear_model.LinearRegression() # 調(diào)用最小二乘回歸方法
model.fit ([[0, 0], [1, 1], [2, 2]], [1, 2, 3]) # 模型擬合
print model.coef_
print model.intercept_
print model.predict([3, 3])

• 決策樹(shù)實(shí)戰(zhàn)項(xiàng)目-鳶尾花分類(lèi)

# -*- coding: utf-8 -*-
from sklearn import datasets
from sklearn import cross_validation
from sklearn import tree
from sklearn import metrics

iris = datasets.load_iris()
iris_X = iris.data
iris_Y = iris.target

feature_train, feature_test, target_train, target_test = 
cross_validation.train_test_split(iris_X, iris_Y,test_size = 0.33,
random_state = 42)

# 所以參數(shù)均置為默認(rèn)狀態(tài)
dt_model = tree.DecisionTreeClassifier()
dt_model.fit(feature_train, target_train)
predict_result = dt_model.predict(feature_test)

print predictpredict_result
# 評(píng)估計(jì)算方法查看預(yù)測(cè)結(jié)果的準(zhǔn)確度
print metrics.accuracy_score(predict_result,target_test)

#采用model自帶的評(píng)估函數(shù)
scores = dt_model.score(feature_test, target_test)
print scores

• Kmeans

$ wget http://labfile.oss.aliyuncs.com/courses/880/cluster_data.csv 獲取數(shù)據(jù)集

import pandas as pd  # 導(dǎo)入數(shù)據(jù)處理模塊
import matplotlib.pyplot as plt #導(dǎo)入繪圖模塊
from sklearn import cluster

# 導(dǎo)入數(shù)據(jù)文件
file = pd.read_csv("cluster_data.csv", header = 0)
print file
X = file['x'] #定義橫坐標(biāo)
y = file['y'] #定義縱坐標(biāo)

#plt.scatter(X,y) #繪制散列點(diǎn)
#plt.show() #顯示圖

kn_model = cluster.k_means(file,n_clusters = 3)
cluster_centers = model[0] # 聚類(lèi)中心數(shù)組
cluster_labels = model[1] # 聚類(lèi)標(biāo)簽數(shù)組
plt.scatter(X, y, c=cluster_labels) # 繪制樣本并按聚類(lèi)標(biāo)簽標(biāo)注顏色
# 繪制聚類(lèi)中心點(diǎn)，標(biāo)記成五角星樣式，以及紅色邊框
for center in cluster_centers:
    plt.scatter(center[0], center[1], marker="p", edgecolors="red")
plt.show() # 顯示圖

參考文獻(xiàn)
1.sklearn 官網(wǎng)