Foundations of AI Trading

This unit introduces the technical foundations for generative AI applications in trading and asset management, covering data acquisition, preprocessing, performance measurement, and development environments (Python/Matlab). Students will learn basic supervised and unsupervised methods, how to translate strategy specifications into backtest code, and core generative model concepts (including diffusion models and representation complexity) with a hands-on sentiment analysis case using large language models. The unit prepares learners to design, test, and evaluate simple AI-driven trading ideas in subsequent units.

Learning Objectives

• Apply programming tools to retrieve and preprocess historical market data and compute adjusted closing prices, daily returns, and performance metrics such as the Sharpe ratio using Python or Matlab
• Demonstrate the ability to translate a trading strategy specification into executable code and construct a basic backtest to evaluate strategy performance
• Analyze supervised and unsupervised learning techniques and evaluate their suitability for market tasks such as signal generation, clustering, and feature extraction
• Demonstrate understanding of generative model concepts (including diffusion models and representation complexity) and apply large language models to perform sentiment analysis on central bank communications