Robust Resource Bounds with Static Analysis and Bayesian Inference

on YouTube ↗20m

" summary

Claims a combined static/data technique that actually implements and evaluates a hybrid: it integrates automatic amortized resource analysis (AARA) with Bayesian inference, uses reflective HMC sampling plus linear-program solving to turn posterior samples into sound typing constraints, and proves Hybrid AARA statistically sound under standard measurement assumptions. Implementation runs on 10 benchmarks that conventional AARA fails, and empirical results show higher proportions of sound bounds and better accuracy than optimization-only or purely data-driven methods. Artifacts are available.

AI-written summary · about

" chapters (13)

1. 0:00Introduction
2. 1:07Static vs data-driven analysis
3. 2:29Static analysis techniques
4. 3:35Data analysis workflow
5. 5:01Bayesian inference for bounds
6. 6:25Worst-case cost and coefficients
7. 8:14Hybrid resource analysis design
8. 10:40Amortized analysis in hybrid
9. 11:23Hybrid analysis architectures
10. 13:49Hybrid analysis on quicksort
11. 15:54Benchmark results
12. 16:57Key takeaways
13. 17:38Q&A