0

I'm trying to figure out what a scheme is called - what is the accuracy metric for the number of true positives out of the top n items, listed descended by probability, where n is the expected # in the population?

For example, if we expect their to be 1000 true cases in a population, and we then put forth our top 1000 items scored by probability, and get 600 correct, the "accuracy" here is 60%. What is this correct term for this "accuracy"?

(I'd want to call it "accuracy at expected inclusion" or something...)

dashnick
  • 159
  • 8
  • 1
    [Why is accuracy not the best measure for assessing classification models?](https://stats.stackexchange.com/q/312780/1352) [Is accuracy an improper scoring rule in a binary classification setting?](https://stats.stackexchange.com/q/359909/1352) [Classification probability threshold](https://stats.stackexchange.com/q/312119/1352) The same problems apply to your KPI, and indeed to all evaluation metrics that rely on hard classifications. Instead, use probabilistic classifications, and evaluate these using [proper scoring rules](https://stats.stackexchange.com/tags/scoring-rules/info). – Stephan Kolassa Apr 14 '21 at 18:40
  • 1
    FWIW, I'm not advocating this method.. I just want to know if it has a name – dashnick Apr 14 '21 at 18:42

0 Answers0