User loginNavigation |
Lifted inference: normalizing loops by evaluation
Lifted inference: normalizing loops by evaluation. Oleg Kiselyov and Chung-chieh Shan. 2009 Workshop on Normalization by Evaluation.
Many loops in probabilistic inference map almost every individual in their domain to the same result. Running such loops symbolically takes time sublinear in the domain size. Using normalization by evaluation with first-class delimited continuations, we lift inference procedures to reap this speed-up without interpretive overhead. To express nested loops, we use multiple control delimiters for metacircular interpretation. To express loops over a powerset domain, we convert nested loops over a subset to unnested loops. The paper is a bit hard to follow, but there are enough little tricks here to merit attentive reading. Or better yet, read the code. The basic PLT idea might be summed as doing abstract interpretation on a shallowly embedded DSL using delimited continuations. By Ehud Lamm at 2009-08-14 02:34 | DSL | Functional | Implementation | Meta-Programming | other blogs | 9811 reads
|
Browse archivesActive forum topics |
Recent comments
6 weeks 4 days ago
6 weeks 6 days ago
7 weeks 13 hours ago
14 weeks 10 hours ago
19 weeks 5 days ago
19 weeks 6 days ago
20 weeks 5 days ago
23 weeks 3 days ago
24 weeks 6 days ago
25 weeks 1 hour ago