Reference · GlossaryThresholdA **cutoff score** that turns a continuous model output into a yes/no (or class) decision.On this pageWhen to useWhen not toExample#When to useTuning precision vs recall — raise threshold to approve fewer high-risk cases.#When not toWhen the model already outputs calibrated class probabilities and you use argmax instead.#Example`pred = 1 if score >= 0.7 else 0` — stricter than 0.5, fewer false approvals.Learn by doing → Learn by doing →Related termsClassificationLoss function← All terms