fengdu78 / Coursera-ML-AndrewNg-Notes

吴恩达老师的机器学习课程个人笔记
32.01k stars 10.63k forks source link

ML-Exercise8中代码思路有错误 #47

Open lyx599 opened 5 years ago

lyx599 commented 5 years ago

image 按照视频中的算法,算出概率后应该再按行进行相乘算出每个样例的概率。之后再进行阈值选择,正确做出来后右上那个点也应该标记为红。 https://github.com/lyx599/ML-AndrewNg-homework/blob/master/ex8_anomaly_detection_and_recommendation/anomaly_detection.ipynb 这是我自己实现的

xuewengeophysics commented 4 years ago

@lyx599 我认为您的思路是对的。谢谢!

icekop commented 2 years ago

同意,源代码用的应该是原始模型,算出概率之后需要相乘。不改代码的话,最后标记出的红色点貌似有错误

yanyiting commented 2 years ago

 收到,谢谢!

sullinger commented 2 years ago

哒啦哒哒哒,收到啦!ps.这里是HFer_杨少

Enya191113 commented 2 years ago

您好!      已经收到您的邮件,我会尽快处理。祝身体健康,万事胜意!