Model Performance
This is where you see how smart our engine is getting. Each model's accuracy is tracked over time. The ensemble combines all models — giving more weight to whichever model is making the most money. Watch these numbers improve as the engine learns from its trades.
—
Awaiting first grading cycle
The percentage of predictions this model got right. 60%+ is considered good for stock prediction — even a small edge over 50% can be very profitable.
Accuracy Over Time
Ensemble Weights How much influence each model has on the final prediction. Models that make more money in paper trading get higher weights. These adjust automatically every day.
Ensemble
Active
● Learning Journal
No journal entries yet. The learning journal will be populated after the first grading cycle completes.