From bb3ed3b0c2dce360b04cf0b8a5cab7cff87b2323 Mon Sep 17 00:00:00 2001 From: 0urobor0s <0urobor0s@users.noreply.github.com> Date: Wed, 18 Oct 2023 17:29:34 +0100 Subject: [PATCH] Improve fbeta_score documentation --- lib/scholar/metrics/classification.ex | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/lib/scholar/metrics/classification.ex b/lib/scholar/metrics/classification.ex index 373c1b45..797d1cd6 100644 --- a/lib/scholar/metrics/classification.ex +++ b/lib/scholar/metrics/classification.ex @@ -629,6 +629,10 @@ defmodule Scholar.Metrics.Classification do If all examples are true negatives, then the result is 0 to avoid zero division. + #{~S''' + $$F_\beta = (1 + \beta^2) \cdot \frac{\mathrm{precision} \cdot \mathrm{recall}}{(\beta^2 \cdot \mathrm{precision}) + \mathrm{recall}}$$ + '''} + ## Options #{NimbleOptions.docs(@fbeta_score_schema)}