XGBoostVolatilityModel
Gradient-boosted tree regression on the log of next-bar realized variance.
Trained artifact not included in repository
models/volatility.py · private research repository
Console
Direction, magnitude and volatility forecasts from the production triad, served through the Voltacent API.
Model inventory
The production triad is available through the Voltacent API in private beta. Research catalog metrics remain dated historical backtest results, not live performance.
Direction classifier
LiveXGBoost classifier
Live in production · private inference cluster
Magnitude regressor
LiveGen-1 XGBoost regressor
Live in production · private inference cluster
Volatility forecaster
LiveGen-1 XGBoost
Live in production · private inference cluster
Production deployment runs all three legs — direction, magnitude gate, volatility gate — as one decision pipeline.
Research catalog — API candidates
Research results · historical backtestxgboost_20260903_200844
ACTIVEGen-1 classifier
lightgbm_20260908_170458
VALIDATED / INACTIVEClassifier
xgboost_reg_20260912_160548
VALIDATED / INACTIVEHybrid regression
xgboost_20260916_103341
STAGINGHybrid classifier
Threshold sweep pending · not LEAN-tested
76 trained fold-models across 5 families in the research registry. Two DL-only ablations failed; one volatility run pending training — not offered. All catalog metrics are dated historical research/backtest results — not live performance.
Live status
This panel queries the production API health endpoint on page load and shows exactly what it returns — nothing cached, nothing assumed.
Contacting health endpoint…
The health endpoint reports whether each model leg of the production triad is loaded and serving. It carries no signals, forecasts, or performance data.
Console
Production inference is available via the private-beta API, separate from the staging candidate and dated canary telemetry snapshot. Historical research results are not inserted into live fields.
Active staging candidate
xgboost_20260916_103341
Direction
Available via API
Served through the Voltacent API — private beta, request access.
Expected return
Available via API
Served through the Voltacent API — private beta, request access.
Volatility
Available via API
Served through the Voltacent API — private beta, request access.
Model health
Available via API
Served through the Voltacent API — private beta, request access.
Data freshness
Available via API
Served through the Voltacent API — private beta, request access.
Last inference
Available via API
Served through the Voltacent API — private beta, request access.
Telemetry: RECORDED CANARY SNAPSHOT · Last recorded: 2026-09-16 08:45 UTC · Streaming: NOT CONNECTED · LEAN: NOT YET TESTED. This is not a live trading dashboard.
Volatility model family · experiment/xgb-volatility-h1
Every variant targets log_realized_variance for the next H1 bar. The baselines exist to show whether XGBoostVolatilityModel adds anything beyond simple persistence and HAR-RV structure.
3 of 5 reported variants verified. Five volatility model types were reported for this branch. Only three are verifiable from the supplied repository facts; the remaining two are not shown because their names and definitions could not be confirmed.
Gradient-boosted tree regression on the log of next-bar realized variance.
Trained artifact not included in repository
models/volatility.py · private research repository
Leakage-safe persistence: forecasts next-H1 variance from the most recent completed realized variance.
Trained artifact not included in repository
models/volatility_baselines.py · private research repository
Heterogeneous autoregressive realized-variance model combining short, medium and long causal windows.
Trained artifact not included in repository
models/volatility_baselines.py · private research repository
Canary telemetry snapshot
A historical canary snapshot recorded in Voltacent telemetry.json. Values describe that recorded observation only and do not stream or imply current status.