In-sample (IS) data is the portion of the historical record on which the strategy's parameters were optimized. The strategy has implicitly memorized the in-sample period; its IS performance is therefore an upper bound on what to expect live, not a forecast.
A wide gap between IS and OOS performance is the classic signature of overfitting: the parameters captured noise specific to the IS period rather than generalizable structure. Healthy strategies degrade modestly from IS to OOS; overfit strategies degrade dramatically.
The IS / OOS split ratio is typically 60–70% IS / 30–40% OOS for a static split, or rolling windows of similar relative size for walk-forward. The exact ratio matters less than the discipline of treating OOS as untouchable.
Cómo Noon Barbari usa In-sample
Cada concepto aquí está implementado en la plataforma. Abre la documentación o la herramienta correspondiente para verlo en acción.
IS / OOS in the platform →Términos relacionados
- Backtesting
Out-of-sample
Data the strategy was not allowed to see during parameter selection.
- Backtesting
Walk-forward optimization
Optimize on a rolling in-sample window, validate on the next out-of-sample slice.
- Backtesting
Overfitting
Fitting a strategy so closely to past data that it captures noise, not signal.
- Backtesting
Backtest
Simulating a trading rule on historical data to estimate how it would have performed.