Topic: Question about OOS

Hi, does anyone know what implications does the OOS have on optimization?
Basically what does it mean?

Re: Question about OOS

“Out of sample” testing is a technique for preventing over-optimization (curve-fitting). The idea is simple: the testing program (Optimizer or Generator) uses a part of the data when searching the best result. The program calculates the backtest for the rest of the data but doesn't consider the end result.

We can evaluate a strategy by the chart pattern after the OSS period.

This strategy is OK.

But this is over-optimized.