Iteration budget

A pre-committed maximum on the number of variants you will test against a body of data. Each pass at an out-of-sample segment leaks its specifics into your choices, so a data segment gets burned after repeated mining and stops being a fair test. Fixing the budget before starting keeps an optimization search from degenerating into data-snooping.

First used in Lesson 2.7 · Walk-forward: the honest way to test an optimized system — the lesson that makes this term real.