Articles

  • Example for recall, precision, acc

    Because time is limited, this is a Chinese article

    本文介绍机器学习里面几个难懂的概念:召回率、精确率、准确率、F-Score、漏检率、虚检率
    由于实习的关系,我先用医疗背景解释这个问题,肿瘤病人设为1,正常人设为0 首先明确几个变量:
    TP:真阳性,有肿瘤的人1被预测为1
    FN:假阴性,有肿瘤的病人1被预测为0,这个比较可怕
    TN:真阴性,没肿瘤的人0被预测为0
    FP:假阳性,没肿瘤的人0被预测为1
    有肿瘤的病人也就是1的总数为P=TP+FN,正常人的总数为N=TN+FP

    Read More »

  • Firstblog

    My first Blog

    Hi, I’m Lui Yi. A geek likes making anything fantastic. Welcome to my blog. I’m trying to learn how to use Github. I appreciate the author in Jekyll gave me the excellent blog template.

    Read More »