Experimenting with a modified SVR algorithm where I've artificially limited the leverage to 2.5 to meet the contest rules. Any advice for better performance? Short backtests generate positive returns but longer backtests always lose money.
Experimenting with a modified SVR algorithm where I've artificially limited the leverage to 2.5 to meet the contest rules. Any advice for better performance? Short backtests generate positive returns but longer backtests always lose money.