Practical AI Recommender systems and high-frequency trading
Mar 23, 2021
David Sweet, author of "Tuning Up: From A/B testing to Bayesian optimization," shares his insights on system tuning and its diverse applications. He explores the fascinating parallels between recommender systems and high-frequency trading, emphasizing the importance of A/B testing and experimentation. Sweet discusses the evolution of trading strategies influenced by AI and offers tips on navigating the challenges of real-world data interpretation. His unique journey from physics to finance provides a lighthearted take on the practicality of AI in trading.
AI Snips
Chapters
Books
Transcript
Episode notes
Path to Finance
- David Sweet got a call from a recruiter and learned about finance jobs for physicists.
- He interviewed at a small company with many computing resources, tasked with building an autonomous trading strategy.
AI in Finance
- Deep learning helps interpret alternative datasets and order books in high-frequency trading.
- Simulation optimization, a classic engineering technique, is crucial for building trading strategies.
Similar Workflows
- The workflow for quantitative traders and recommender system engineers is similar, involving ideation, offline testing, and online experimentation.
- Experimentation is the most accurate but hardest part due to limited data.
