Book Guide

Use this section to move between the production ml4t.diagnostic API and the pedagogical workflows in Machine Learning for Trading, Third Edition.

The book code lives under:

• third_edition/code/<chapter>/... for chapter notebooks/scripts
• third_edition/code/case_studies/<case_study>/... for end-to-end workflows

How To Use This Guide

• Start in the book when you want concepts, derivations, and step-by-step builds.
• Start in the library docs when you want the stable API, configuration patterns, and reusable workflows.
• Use the chapter map below to jump between the two.

Chapter Map

Book chapter	What it teaches	Relevant ml4t.diagnostic APIs	Book paths
Ch06 Strategy Definition	Leak-free time-series validation, purging, embargo, fold visualization	WalkForwardCV, CombinatorialCV, plot_cv_folds()	code/06_strategy_definition/01_cv_foundations.py
Ch07 Defining the Learning Task	Multiple testing, FDR, HAC-adjusted IC, causal sanity checks	benjamini_hochberg_fdr(), compute_ic_hac_stats(), compute_min_trl(), compute_pbo()	code/07_defining_learning_task/07_multiple_testing.py, code/07_defining_learning_task/08_causal_sanity_checks.py
Ch08 Feature Engineering	Feature triage, robustness checks, event studies	analyze_signal(), analyze_ml_importance(), FeatureSelector, EventStudyAnalysis	code/08_feature_engineering/05_feature_selection.py, code/08_feature_engineering/06_robustness_sensitivity.py, code/08_feature_engineering/07_event_studies.py
Ch09 Model-Based Features	Stationarity, distribution diagnostics, autocorrelation, drift	FeatureDiagnostics, analyze_stationarity(), analyze_distribution(), analyze_autocorrelation(), drift tools	code/09_model_based_features/01_visual_diagnostics.py, code/09_model_based_features/07_arima_features.py, code/09_model_based_features/08_garch_volatility.py, code/09_model_based_features/12_wasserstein_regimes.py
Ch16 Strategy Simulation	Performance reporting, tearsheets, Sharpe inference, DSR, RAS	PortfolioAnalysis, BacktestProfile, generate_backtest_tearsheet(), generate_tearsheet_from_result(), deflated_sharpe_ratio(), compute_pbo()	code/16_strategy_simulation/09_performance_reporting.py, code/16_strategy_simulation/11_sharpe_ratio_inference.py, code/16_strategy_simulation/12_dsr_validation.py, code/16_strategy_simulation/13_ras_protocol.py
Ch17 Portfolio Construction	Portfolio metrics, allocator comparison, robust optimization	PortfolioAnalysis, compute_allocator_metrics() wrappers used in the book	code/17_portfolio_construction/01_portfolio_metrics.py, code/17_portfolio_construction/02_mean_variance_optimization.py, code/17_portfolio_construction/03_robust_optimization.py, code/17_portfolio_construction/04_kelly_criterion.py, code/17_portfolio_construction/06_hierarchical_risk_parity.py, code/17_portfolio_construction/08_library_comparison.py, code/17_portfolio_construction/09_allocation_horse_race.py
Ch19 Risk Management	VaR/CVaR, barrier analysis, factor exposure, trade-SHAP, drift	analyze_distribution(), BarrierAnalysis, compute_factor_model(), compute_return_attribution(), compute_risk_attribution(), TradeAnalysis, TradeShapAnalyzer	code/19_risk_management/01_var_cvar.py, code/19_risk_management/02_exit_strategies.py, code/19_risk_management/04_factor_exposure.py, code/19_risk_management/05_trade_shap_diagnostics.py, code/19_risk_management/07_drift_detection.py

From Book Notebook To Production Workflow

The book often builds the method manually first and then introduces the library call:

• Ch06 shows fold arithmetic and purge logic before using WalkForwardCV.
• Ch07 computes statistical adjustments step by step before switching to compute_ic_hac_stats() and benjamini_hochberg_fdr().
• Ch08 explains feature triage manually, then points to analyze_signal() and FeatureSelector for reusable pipelines.
• Ch16 and the case studies wrap portfolio and tearsheet reporting behind shared helpers that call PortfolioAnalysis and the backtest visualization layer.

For the end-to-end view, continue to Case Studies.