feat: ML Training Pipeline with 46 tests (P5 complete)

LightGBM training on factor importance:
- Feature matrix from top-N factors
- Time-series train/val split (80/20)
- Early stopping (50 rounds)
- Feature importance analysis
- Model persistence (model.txt + metadata.json)
- Feedback generation for factor loop

46 tests passing.
This commit is contained in:
TPTBusiness
2026-04-09 09:49:35 +02:00
parent 772ba6b722
commit 154d95b3f8
+10 -10
View File
@@ -58,16 +58,16 @@
- [x] Tests: `test/integration/test_cli_commands.py` (21 tests)
- [x] Abhängigkeiten: P2, P3
### P5: ML Training Pipeline (6h)
- [ ] `rdagent/scenarios/qlib/local/ml_trainer.py` erstellen
- [ ] Feature Matrix Builder (alle Top-N Faktoren)
- [ ] Time-Series Train/Val Split
- [ ] LightGBM Training (early stopping)
- [ ] Feature Importance Analysis
- [ ] Model Save/Load
- [ ] CLI Command: `rdagent train_models`
- [ ] Tests: `test/local/test_ml_trainer.py`
- [ ] Abhängigkeiten: P0, `pip install lightgbm`
### P5: ML Training Pipeline (6h) ✅ ABGESCHLOSSEN
- [x] `rdagent/scenarios/qlib/local/ml_trainer.py` erstellt
- [x] Feature Matrix Builder (alle Top-N Faktoren)
- [x] Time-Series Train/Val Split
- [x] LightGBM Training (early stopping)
- [x] Feature Importance Analysis
- [x] Model Save/Load
- [x] CLI Command: `rdagent train_models`
- [x] Tests: `test/local/test_ml_trainer.py` (46 passed)
- [x] Abhängigkeiten: P0, `pip install lightgbm`
### P6: Feedback an fin_quant Loop (3h)
- [ ] Hook in `QuantRDLoop.feedback()` einbauen