Recall rate and Accuracy

Recall rate and accuracy are frequently two concepts and indicators involved in data mining prediction and search engines on the Internet.
Recall rate: Recall, also known as "Check full rate"-or a good record of check full rate, can also better reflect the actual quality.
Accuracy: precision, also known as "precision" and "accuracy ".

Taking retrieval as an example, you can use the following information to describe the search:

	Related	Unrelated
Retrieved	A	B
Not retrieved	C	D

 

 

A: The retrieved and related              (Search for what you want)
B: retrieved, but unrelated        (Found but useless)
C: not retrieved, but related      (Not found, but actually wanted)
D: unretrieved and irrelevant        (It is useless if no search is found)

If we want to: the more content we retrieve, the better. This is the pursuit of "query full rate", that is, a/(A + C). The larger the value, the better.
If we want to: In the retrieved documents, what we really want, that is, the more relevant the better, the less irrelevant the better, this is the pursuit of "accuracy ", that is, a/(A + B), the bigger the better.
 
Although there is no inevitable relationship between "recall rate" and "accuracy rate" (as can be seen from the above formula), they are mutually restricted in actual application. Find a balance point based on actual needs.

It is often difficult to quickly respond to the "recall rate ". I think this is also related to the literal meaning, and the meaning cannot be directly seen from the literal meaning of "Recall. The meaning of "Recall" in Chinese is to call XX back. The English "Recall" corresponding to the "recall rate". In addition to the meaning of "order something to return" mentioned above, the recall also refers to "remember.
Recall: the ability to remember something. That you have learned or something. That has happened in the past.
When we query all the details of a System Event (enter the query term), recall refers to the number of details that the system can "Recall" those events, in general, it is "the ability to recall ". "Number of details that can be recalled" divided by "all details that the system knows about this matter", that is, "memory rate", that is, recall-recall rate. It is simple and can also be understood as the full query rate.