模型评估

# -*- coding: utf-8 -*-
"""
Created on Thu Sep 27 16:24:29 2018
    模型及预测准确度评估
@author: zhen
"""

from sklearn import metrics

if __name__ == "__main__":
    
    # 同一性homogeneity：每个群集只包含单个类的成员。 
    # 完整性completeness：给定类的所有成员都分配给同一个群集。
    # 调和平均V-measure
    y = [0, 0, 0, 1, 1, 1]
    y_hat = [0, 0, 1, 1, 2, 2]
    h = metrics.homogeneity_score(y, y_hat)
    c = metrics.completeness_score(y, y_hat)
        
    v2 = 2 * c * h / (c + h)
    v = metrics.v_measure_score(y, y_hat)
    print(u'同一性(Homogeneity):', h)
    print(u'完整性(Completeness):', c)
    print(u'V_Measure:', v2, v)
    
    y = [0, 0, 0, 1, 1, 1]
    y_hat = [0, 0, 1, 3, 3, 3]
    h = metrics.homogeneity_score(y, y_hat)
    c = metrics.completeness_score(y, y_hat)
    v = metrics.v_measure_score(y, y_hat)
    
    print(u'同一性(Homogeneity):', h)
    print(u'完整性(Completeness):', c)
    print(u'V_Measure:', v)
    
    # 允许不同值
    y = [0, 0, 0, 1, 1, 1]
    y_hat = [1, 1, 1, 0, 0, 0]
    h = metrics.homogeneity_score(y, y_hat)
    c = metrics.completeness_score(y, y_hat)
    v = metrics.v_measure_score(y, y_hat)
    
    print(u'同一性(Homogeneity):', h)
    print(u'完整性(Completeness):', c)
    print(u'V_Measure:', v)
    
    # 兰德指数
    # ARI值的范围是[-1,1]，负的结果都是较差的，越接近-1表示聚类越差，
    # 正的结果都是较好的，1是最佳结果，越接近1表示聚类结果越好。
    y = [0, 0, 1, 1]
    y_hat = [0, 1, 0, 1]
    ari = metrics.adjusted_rand_score(y, y_hat)
    print("兰德指数:", ari)
    
    y = [0, 0, 0, 1, 1, 1]
    y_hat = [0, 0, 1, 1, 2, 2]
    ari = metrics.adjusted_rand_score(y, y_hat)
    print("兰德指数:", ari)
    

结果：