0.25
0.5
0.75
1.25
1.5
1.75
2
Lightweight Implementations of Probabilistic Programming Languages Via Transformational Compilation
Published on May 06, 20116421 Views
We describe a general method of transforming arbitrary programming languages into probabilistic programming languages with straightforward MCMC inference engines. Random choices in the program are
Related categories
Chapter list
Lightweight Implementations of Probabilistic Programming Languages via Transformational Compilation00:00
Implementing Probabilistic Programming LanguagesWithout the Agonizing Pain00:14
Summary00:20
Outline01:17
Probabilistic Programming01:34
The Big Idea01:42
Distributions over Traces02:21
Example: LDA03:34
Nonparametrics04:06
Example: ((H)DP)MM05:20
Meta-Modeling07:03
Lightweight PPL Implementations07:38
Inference07:43
Observation: Execution Trace (1)08:09
Observation: Execution Trace (2)09:18
Transformational Compilation09:38
MCMC over Execution Traces10:09
But What Name? (1)11:21
But What Name? (2)12:31
But What Name? (3)12:53
Generating Names13:44
Example: Geometric15:19
Minimal Interpretative Overhead15:58
New Inference Options17:25
Different Inference Options17:39
Dynamic Dependency Analysis (1)19:08
Dynamic Dependency Analysis (2)20:11
Dynamic Dependency Analysis (3)20:31
Example: mesh inference21:18
Summary22:41
Thank you!23:31