[学习笔记]机器学习模型结果分析与检验

最新推荐文章于 2024-04-12 15:15:21 发布

强heaven

最新推荐文章于 2024-04-12 15:15:21 发布

阅读量853

点赞数 1

分类专栏： AI机器学习文章标签：机器学习

本文链接：https://blog.csdn.net/qq_25888559/article/details/120638288

版权

AI机器学习专栏收录该内容

5 篇文章 1 订阅

订阅专栏

一、回归评价指标

# -*- coding: utf-8 -*-

"""
Created on Fri Jun 18 09:41:40 2021
@author: zhuchunqiang
"""

from sklearn.metrics import mean_squared_error,mean_absolute_error,r2_score
y_true = [1,2,4]
y_pred = [1,3,5]

RMSE = mean_squared_error(y_true,y_pred)**0.5
print('均方根误差: %.5f' % RMSE)

MSE = mean_squared_error(y_true,y_pred)
print('均方误差: %.5f' % MSE)

MAE = mean_absolute_error(y_true,y_pred)
print('平均绝对误差: %.5f' % MAE)

R2 = r2_score(y_true,y_pred)
print('R2: %.5f' % R2)

二、分类评价指标

# -*- coding: utf-8 -*-
"""
Created on Fri Jun 18 09:51:15 2021
@author: zhuchunqiang
"""

from sklearn.metrics import precision_score,recall_score,accuracy_score,f1_score
y_true = [3, 1, 2, 0, 1, 0]
y_pred = [1, 0, 3, 0, 2, 1]

prec    =   precision_score(y_true, y_pred, average='micro')
print('分类精度: %.5f' % prec)

recall  =   recall_score(y_true, y_pred, average='micro')
print('召回率: %.5f' % recall)

acc     =   accuracy_score(y_true, y_pred)
print('准确率: %.5f' % acc)

f1      =   f1_score(y_true, y_pred, average='micro')
print('调和平均f1: %.5f' % f1)