Predicting Query Execution Time for JIT Compiled Database Engines

This lecture discusses the challenges of accurately predicting query execution time for Just-In-Time (JIT) compiled database engines. The instructor presents a baseline analytical model and introduces a JIT prediction approach using active learning. Experimental results on an Intel Xeon system with a database engine based on JIT compilation show a significant improvement over the baseline model. The lecture concludes by highlighting the limitations of analytical models and the potential of machine learning techniques in query time prediction.