Jarvis V2

v2.0.0
0/7 Agents
Uptime:
BQ
API
Live
Loading market data...

ML Model Fleet

Planned ML model fleet — backtesting metrics from research phase

ML Pipeline — Development Roadmap

These models are planned capabilities. Metrics shown are from backtesting research, not live trading. The ensemble system will integrate with Trinity consensus once deployed.

Model Fleet Overview

Active

Random Forest

Accuracy

62.3%

Predictions14,832
Last TrainedFeb 14, 2026
Sharpe1.42
Ensemble Weight22.0%
ActiveBest

Gradient Boosting (XGBoost)

Accuracy

65.7%

Predictions14,832
Last TrainedFeb 14, 2026
Sharpe1.67
Ensemble Weight30.0%
Active

Ridge Regression

Accuracy

58.1%

Predictions14,832
Last TrainedFeb 14, 2026
Sharpe1.11
Ensemble Weight13.0%
Training

LSTM Neural Network

Accuracy

64.1%

Predictions12,416
Last TrainedFeb 13, 2026
Sharpe1.53
Ensemble Weight25.0%
Stale

Support Vector Regression

Accuracy

59.2%

Predictions11,204
Last TrainedFeb 11, 2026
Sharpe1.19
Ensemble Weight10.0%

Performance Comparison

ModelAccuracyPrecisionRecallF1 ScoreAUC-ROCSharpe Ratio
RF
62.3%64.1%59.8%61.9%67.1%1.42
XGB
65.7%67.2%63.1%65.1%70.9%1.67
RIDGE
58.1%59.4%56.7%58.0%62.3%1.11
LSTM
64.1%65.8%61.2%63.4%69.4%1.53
SVR
59.2%60.8%57.4%59.1%63.8%1.19

Feature Engineering

FeatureCategoryRFXGBRIDGELSTMSVR
Price MomentumTechnical+++++
Volume ProfileTechnical+++++
Sentiment ScoreAlternative++-++
Options FlowDerivatives++-+-
Insider ActivityFundamental++--+
Macro IndicatorsMacro+++++
Sector RotationMacro+++--
Earnings Surprise HistoryFundamental++++-
Total Features8/88/85/86/85/8

Backtesting Results

ModelTotal TradesWin RateAvg Return/TradeMax DrawdownProfit FactorCalmar Ratio
RF54160.1%+0.41%-6.8%1.381.22
XGB54164.3%+0.47%-5.9%1.521.41
RIDGE54156.8%+0.27%-7.4%1.220.89
LSTM46862.2%+0.43%-6.4%1.441.28
SVR41357.6%+0.29%-8.2%1.250.91

Model Ensemble

Weighted average ensemble — final signal is a confidence-weighted combination of all active model predictions.

RF
22.0%
XGB
30.0%
RIDGE
13.0%
LSTM
25.0%
SVR
10.0%

Signal Calculation

S_final = sum(w_i * p_i) / sum(w_i)

where w_i = model weight, p_i = model prediction

Current Ensemble Confidence64.8%
Active Models3/5
Signal DirectionBULLISH

Training Pipeline

Last Complete Run:Feb 14, 2026 — 8:35 AM

Data Collection

Completed Feb 14, 8:00 AM

4m 12s

Feature Engineering

Completed Feb 14, 8:04 AM

7m 38s

Model Training

In progress...

18m 41s

Validation

Waiting

--

Deployment

Waiting

--
ScheduleEvery 6 hours
Next Scheduled RunFeb 14, 2026 — 2:00 PM
Pipeline StatusTraining In Progress
GPU Utilization78.4%
LSTM Neural Network — Retraining with updated feature set (epoch 847/1200)