|
|
|
|
|
|
by adjkant
2934 days ago
|
|
|
No other baselines (I'm only going for practical not theoretical or publishing so no need really). Just market features but I'm sure other features are out there. At the minutes level, predicting crashes doesn't do much good as often crashes have ideal 1 hour buying windows so the macro crash/spikes don't really matter, and if they did, would be caputed by the model still. |
|
|