docs(book): Reformat chapters 13-15 with rich markdown

Continue enhancing book chapters with visual formatting improvements.

Chapter 13 (AI Sentiment Trading):
- 15+ callout boxes for key concepts, empirical results, warnings
- 7 Mermaid diagrams (sentiment processing, BERT architecture, fusion workflow)
- 12+ comparison tables (dictionaries, NLP approaches, performance metrics)
- Enhanced code blocks with detailed explanations

Chapter 14 (ML Prediction):
- 10+ callout boxes highlighting overfitting prevention and best practices
- 3 Mermaid diagrams (ML evolution, pipeline, LSTM architecture)
- 5+ tables (indicators, hyperparameters, regime detection, cost analysis)
- Better section hierarchy with granular subheadings

Chapter 15 (PumpSwap/MEV):
- 8+ callout boxes for DeFi-specific risks and opportunities
- 7 Mermaid diagrams (MEV extraction flows, Jito architecture, timelines)
- 15+ tables (MEV economics, strategies, latency comparison, scam detection)
- Enhanced OVSM code with anti-rug checks and gas optimization

All chapters now include:
- Visual callouts (💡📊⚠️💻🎯🔬)
- Algorithm flowcharts and sequence diagrams
- Comparison tables for methods and metrics
- Horizontal rules for clear section breaks
- Improved heading hierarchy for better navigation

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <[email protected]>

Author: 0xrinegade