What do you understand by Recall and Precision?
Precision and recall both are metrics used for understanding the performance of your classification algorithm.
In simple words
Precision: It tells you how many are records are classified correctly.
Recall: It tells you how well your algorithm is able to study the pattern in your data and upto what extent it can recall that.
Suppose in a village of 100 people, 30 people got infested with a rare virus, which starts showing visible signs in a month. You are a clinician and you decided to use a detector to identify infested people before it’s too late . Your device detected that 40 out of 100 people were infested with virus (Detected Positive). Your findings now mapped with actual cases as follows:
Precision and Recall of your detector can be understood by answering following questions.
Recall: What percentage of people actually infested with virus were detected correctly using your device? = 20/ 30 or 66.67%
Precision: What percentage of people detected positive using your device were actually infested with virus? = 20/ 40 or 50.00%