4676fe79e2
- ml-observability (1.2K lines): alerts, dashboards, metrics modules. Depends on ml-core + common (ModelType). 4 tests passing. - ml-stress-testing (1.3K lines): load_generator, market_simulator, performance_analyzer modules. Depends on ml-core + common + config. 5 tests passing. - ml-security (1.4K lines): anomaly_detector, prediction_validator modules. Depends on ml-core + ml-ensemble (EnsembleDecision, ModelVote, TradingAction). 17 tests passing. Total: 18 sub-crates extracted from ml monolith. Workspace: 0 errors, ml tests 876 + 26 in new sub-crates. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
ml
10-model ML ensemble for the Foxhunt HFT system, built on Candle v0.9.1.
Models
- DQN (Rainbow) — deep Q-network with prioritized replay, dueling heads, noisy nets
- PPO — proximal policy optimization with GAE, LSTM policies, clip-higher
- TFT — temporal fusion transformer for multi-horizon forecasting
- Mamba2 — state space model for sequence prediction
- Liquid Networks — biologically inspired networks for non-stationary data
- TLOB — transformer-based limit order book analysis
- KAN — Kolmogorov-Arnold networks
- xLSTM — extended LSTM architecture
- TGGN — temporal graph neural network
- Diffusion — diffusion-based generative model
Key Modules
ensemble— model ensemble coordination and confidence aggregationhyperopt— PSO-based hyperparameter optimization with per-model adapterstrainers— unified training loops (DQN, PPO, supervised)inference—InferenceAdaptertrait for predictioncheckpoint— model checkpointing and restorationevaluation— walk-forward evaluation pipeline
Usage
use ml::dqn::DQN;
use ml::ppo::PpoTrainer;