Overfitting
Definition
When a strategy is too closely tailored to historical data, capturing noise rather than genuine patterns. Overfit strategies perform well in backtests but poorly in live trading.
Example
A strategy with 50 parameters perfectly fit to 6 months of data is likely overfit and won't generalize to future markets.
How Gilito AI Uses It
Walk-forward validation is our primary defense against overfitting. Strategies must prove themselves on out-of-sample data to receive high Gilito Scores.