ToolJoltTools

F-beta Score Calculator

Weighted harmonic mean of precision and recall — tune β to favor recall (β>1) or precision (β<1) for your problem.

F-β score
F1 (for reference)

F2 (β=2) is standard when missing positives is costly (medical screening, fraud recall); F0.5 (β=0.5) when false alarms are costly (spam, recommendation precision). F1 (β=1) weighs them equally.

Formula

Fβ = (1+β²)·P·R / (β²·P + R) — β is how many times more important recall is than precision
References: van Rijsbergen (1979), Information Retrieval, 2nd ed. (E-measure / F-measure)

About F-beta Score Calculator

F1 weighs precision and recall equally — but most real problems don't. The F-beta score generalizes it with a single knob: β is how many times more important recall is than precision. Use F2 when a missed positive (undiagnosed disease, undetected fraud) hurts more than a false alarm; use F0.5 when false alarms (spam in the inbox, irrelevant recommendations) are the costlier mistake. This calculator computes Fβ for any β and shows F1 alongside so you can see exactly how the weighting shifts the score.

How to use F-beta Score Calculator

  1. 1Enter your values into F-beta Score Calculator — sensible, domain-typical defaults are pre-filled so you see a real result immediately.
  2. 2The result recomputes live using the formula shown on the page; there is no button to press.
  3. 3Adjust any input to compare scenarios, then read the worked example to see the substituted numbers.

Why use F-beta Score Calculator?

  • Computes F-beta Score instantly in your browser — no sign-up, no upload, no server round-trip.
  • 100% free and unlimited, with the exact formula shown: Fβ = (1+β²).
  • Runs entirely client-side, so every value you enter stays private on your device.
  • Live recompute as you type, with a worked example and authoritative references for trust.

Frequently asked questions

How do I choose β?+

Set β to the ratio of false-negative cost to false-positive cost. If missing a positive is 4× worse than a false alarm, β=2 (β² ≈ 4) roughly encodes that. It's a heuristic, not exact decision theory, but it aligns the metric with your real priorities far better than defaulting to F1.

Why a harmonic mean instead of an average?+

The harmonic mean punishes imbalance: a model with precision 1.0 and recall 0.0 has an arithmetic mean of 0.5 but an F-score of 0. This forces both quantities to be reasonable, preventing the metric from rewarding a model that's excellent at one and useless at the other.

What are F2 and F0.5 exactly?+

F2 (β=2) weights recall four times as heavily as precision in the denominator — the screening/fraud default. F0.5 (β=0.5) weights precision four times as heavily — the spam/recommendation default. Both are just Fβ with specific β values, computed identically by this tool.

Can Fβ exceed 1 or go negative?+

No — like precision and recall, it's bounded in [0, 1], where 1 means perfect on both. (That distinguishes it from MCC, which spans −1 to +1.) If you're seeing values outside this range, your precision or recall inputs are outside [0, 1].

Related tools

Related ML & AI tools

Sponsored