AI Prediction Engine
Prophet-based time-series forecasting for key indicators
Training Prophet Model...
Model Methodology
The MacroNode Prediction Engine utilizes Meta's Prophet Engine, an additive model designed for forecasting time-series data where non-linear trends are fit with yearly, weekly, and daily seasonality.
Growth Trend
Fits piecewise linear or logistic growth curves, automatically detecting "changepoints" in economic momentum.
Seasonality
Uses Fourier series to provide a partial-period fit to the yearly and quarterly economic cycles.
Confidence Interval
MAP estimation provides the uncertainty interval (yhat_upper/lower) based on historical variance.
Holiday Effects
Accounts for predictable shocks and outlier events (like COVID-19 or stimulus shifts).
Institutional Outlook
High-fidelity trend projection with error-correction enabled.
*Prophet version 1.1.5 with Stan-backend optimization.