Deep Learning for Financial Market Microstructure

Aim

Workshop Objectives

Workshop Structure

📅 Day 1 — Hands-On Market Microstructure Data Engineering

• Focus: Practical processing of high-frequency tick and order book data
• Hands-On Activities:
  • Importing and cleaning tick-by-tick NASDAQ / crypto order book data
  • Constructing order flow imbalance and liquidity features
  • Generating volatility signature plots
  • Handling asynchronous multi-asset data streams
  • Building efficient data pipelines for large-scale datasets

📅 Day 2 — Hands-On Deep Learning for Financial Time Series

• Focus: Building predictive and execution models
• Hands-On Activities:
  • Developing LSTM-Attention models for price direction prediction
  • Implementing Transformer architectures for volatility forecasting
  • Training and evaluating deep learning models using PyTorch
  • Designing reinforcement learning agents for optimal execution
  • Performance tuning and model validation

📅 Day 3 — Hands-On Backtesting & Strategy Evaluation

• Focus: Strategy validation and research-ready outputs
• Hands-On Activities:
  • Implementing vectorized backtesting with transaction cost modeling
  • Calculating Sharpe ratios, drawdowns, and risk-adjusted metrics
  • Generating performance heatmaps and regime analysis
  • Exporting LaTeX-formatted performance tables
  • Building a complete reproducible quant research pipeline

Who Should Enrol?

Important Dates

Workshop Outcomes