Unbiased (model free), near real-time Machine Learned ensemble forecasts for time-series.
Using customer’s historical time-series, an ensemble forecast for markets is generated in a manner very similar to numerical weather prediction. Instead of making a single forecast using Precision Alpha’s exact mathematical probabilities, a set of 100 forecasts in automatically produced. The ensemble of 100 forecasts indicates the range of possible future states. Elevated thermal probabilities due to non-equilibrium and the low-dimensionality of the problem, make for very informative predictions.
Forecast subscribers have access to their own Precision Alpha S3 sub-bucket. The subscriber loads time-series to the S3 bucket, and the uploaded file is automatically submitted for forecasting. Subscribers pay a monthly subscription and for usage.
Download Below
Closing Prices to Mechanical and Thermal Probabilities
Closing Prices to Single Forecast (30 Day Horizon)
Closing Prices to Ensemble Forecast (100 single forecasts, 10 Day)
Percentage (y-axis) of 100 individual forecasts that reach expected return (x-axis) over next ten trading days.
