From Functional Anlaysis to Reinforcement Learning and personalized search¶

Welcome to the documentation for Reinforcement Learning for Dynamic Search Boost Optimization.

This project combines a synthetic e-commerce search simulator with book-quality documentation exploring RL techniques for search ranking optimization.

Quick Links¶

• Book Overview - Start here for the textbook content
• Book Outline - Chapter-to-code mapping
• Knowledge Graph - Concept mapping and validation tools

Book Chapters¶

Part I - Foundations¶

• Chapter 0: Motivation & First Experiment
• Chapter 1: Search Ranking as Optimization
• Chapter 2: Probability, Measure, and Click Models
• Chapter 3: Stochastic Processes & Bellman Foundations

Part II - Simulator¶

• Chapter 4: Generative World Design
• Chapter 5: Relevance, Features & Reward

Part III - Policies¶

• Chapter 6: Discrete Template Bandits
• Chapter 6a: Neural Bandits
• Chapter 7: Continuous Actions
• Chapter 8: Policy Gradients

Part IV - Evaluation & Deployment¶

• Chapter 9: Off-Policy Evaluation
• Chapter 10: Robustness & Guardrails

Getting Started¶

# Create and activate virtual environment
python -m venv .venv && source .venv/bin/activate

# Install project in editable mode
python -m pip install -e .[dev]

# Run tests
pytest -q