feat(QC): ML-based projects from HandsOnAITrading book + official docs

[jsboige]

Context

We have the HandsOnAITrading book repo as inspiration and QC official docs with ML examples. Our course notebooks should improve in synergy with the QC projects we iterate on.

Objective

Create more serious ML-based trading projects that go beyond simple EMA crossings:

1. Map book examples to our series (extends docs(QC): map HandsOnAITradingBook examples to our series #107)
2. Implement selected examples as QC projects with research notebooks
3. Leverage official QC ML docs: feature engineering, model training, walk-forward
4. Synergy loop: Course notebooks teach the concepts, QC projects demonstrate real application

Specific projects to explore

• Feature engineering: Technical indicators + fundamental data as ML features
• Classification models: Market regime prediction (bull/bear/sideways)
• Time series models: LSTM/Transformer for price prediction
• Reinforcement learning: RL-based portfolio allocation
• Ensemble methods: Combining multiple weak signals

Deliverables

• Research notebooks (QuantBook) for each approach
• Deployable algorithms (main.py) for successful approaches
• Course notebook improvements reflecting the ML patterns used

Related issues

• docs(QC): map HandsOnAITradingBook examples to our series #107: Map HandsOnAITradingBook examples
• feat(QC): create companion QuantBook research notebooks for ML series #108: QuantBook research notebooks
• feat(QC): create deployable algorithms for ML notebook series #109: Deployable algorithms
• QC: Combler les gaps entre projets et notebooks pedagogiques #47: Gaps between projects and notebooks

[jsboige]

Update 2026-03-22 - État d'avancement

Projets ML implémentés (13 projets)

Inspirés du HandsOnAITradingBook et de la doc officielle QC:

Projet	main.py	research.ipynb	Source
ML-Classification	✅	❌	Book Ch06
ML-Regression	✅	❌	Book Ch06
ML-RandomForest	✅	❌	Book Ch06
ML-SVM	✅	❌	Book Ch06
ML-XGBoost	✅	❌	Book Ch06
ML-Ensemble	✅	❌	Book Ch06
ML-DeepLearning	✅	❌	Book Ch06
ML-TextClassification	✅	❌	Book Ch06
ML-FeatureEngineering	❌	❌	Book Ch06
ML-EnhancedPairs	✅	❌	Book Ch06+08
DL-LSTM	✅	❌	Book Ch06
RL-Portfolio	✅	❌	Book Ch07
BTC-ML	✅	✅	Original

Ce qui manque

• BOOK_MAPPING.md (docs(QC): map HandsOnAITradingBook examples to our series #107) : pas encore créé
• Research notebooks : seul BTC-ML en a un, les 12 autres n'ont que main.py
• READMEs : aucun des ML projects n'a de README dans le dépôt (ceux de po-2024 étaient sur une branche non mergée)
• Couverture livre : Ch06 bien couvert (~12/19 exemples), Ch07 partiellement (RL sans options hedging), Ch08 faiblement (pas de PCA+regime dédié)

Prochaines étapes

1. Créer BOOK_MAPPING.md pour documenter la correspondance (docs(QC): map HandsOnAITradingBook examples to our series #107)
2. Ajouter research notebooks pour les projets ML les plus pédagogiques
3. Couvrir les lacunes Ch07 (RL hedging) et Ch08 (PCA+regime, portfolio optimization)

Contexte Jared

Jared recommande explicitement l'utilisation du livre (email 22 mars 2026). Le livre est disponible sur https://www.hands-on-ai-trading.com/