Index Tuning with Multi-armed Bandits

This lecture by the instructor covers the use of Multi-armed bandits (MAB) for index tuning in databases, focusing on the challenges faced in index tuning and the benefits of using MAB algorithms. The presentation discusses the application of MAB in scenarios with static and ad hoc queries, showcasing its efficiency in comparison to traditional tuning tools. The lecture also explores the limitations of general Reinforcement Learning (RL) in this context and highlights the advantages of using MAB for index optimization. The talk concludes with the potential future applications of MAB in physical design tuning and the opportunities for innovation in database optimization.