简体   繁体   中英

How to draw a precision-recall diagram?

Precision and Recall are some metrics for evaluation algorithms, and they are defined like this:

precision = true_positive / (true_positive + false_positive)

recall = true_positive / (true_positive + false_negative)

So, I think for every algorithm we have one number for precision and one number for recall, but we see that there are some diagram for evaluating according to precision and recall, I want to know that how can we draw a diagram with just one point?(one precision and one recall for every algorithm)

You could draw a histogram with the precision and recall for each algorithm. But it's still a bit difficult to read because for each algorithm there are two values. But you need to decide what is the most important characteristics: a good recall or a good precision?

Have you considered using a F-measure that tries to combine both measures? http://en.wikipedia.org/wiki/F1_score

Edit:

My suggestion was to use a histogram:

直方图

But another solution is to plot like this:

云

Both were done using LibreOffice.

The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM