DaStapo

[DaStapo @Stapoh]

@laurentzeimes You probably still can vectorize your backtest on the batch dimension.

[DaStapo @Stapoh]

@2gay2lift Winter Sleep

[DaStapo @Stapoh]

@laurentzeimes Is this a strategy built upon a ML prediction model?

[DaStapo @Stapoh]

@SimpelAlpha Where did you track that? The FTX API updates it only every 15 seconds and Binance does it at random intervals (about 30s).

[DaStapo @Stapoh]

@cryptoquantHQ Are your features "readable" enough so that you can now look at your winning phenotypes and clearly learn something about the market?

[DaStapo @Stapoh]

@cryptoquantHQ What advantage do you see in this approach compared to decision trees (if I understand correctly, the GE searches for combinations of conditions based on features)?