Incremental Precision-Preserving Symbolic Inference for Probabilistic Programs
We present \textsc{ISymb} an incremental
symbolic inference framework
for probabilistic programs in situations when
some loop-manipulated array data, upon
which their probabilistic models are conditioned,
undergoes small changes. To tackle the path
explosion challenge, \textsc{ISymb} is intra-procedurally
path-sensitive except
that it conducts a ``meet-over-all-paths''
analysis within an iteration of a loop
(conditioned on some observed array data).
By recomputing only the
probability distributions for the
paths affected, \textsc{ISymb} avoids
expensive symbolic inference from scratch while
also being
precision-preserving. Our evaluation with a set of
existing benchmarks shows that
\textsc{ISymb} can lead to orders of magnitude
performance improvements compared to its
non-incremental counterpart (under small changes in
observed array data).
Mon 24 Jun
14:00 - 15:30: PLDI Research Papers - Probabilistic Programming at 224AB Chair(s): Martin HirzelIBM Research | ||||||||||||||||||||||||||||||||||||||||||
14:00 - 14:20 Talk | Steffen SmolkaCornell University, Praveen KumarCornell University, David M. KahnCarnegie Mellon University, USA, Nate FosterCornell University, Justin HsuUniversity of Wisconsin-Madison, USA, Dexter KozenCornell University, Alexandra SilvaUniversity College London DOI Pre-print Media Attached | |||||||||||||||||||||||||||||||||||||||||
14:20 - 14:40 Talk | Peixin WangShanghai Jiao Tong University, Hongfei FuIST Austria, Amir Kafshdar GoharshadyIST Austria, Krishnendu ChatterjeeIST Austria, Xudong QinEast China Normal University, China, Wenjun ShiEast China Normal University, China Media Attached | |||||||||||||||||||||||||||||||||||||||||
14:40 - 15:00 Talk | Marco Cusumano-TownerMIT-CSAIL, Feras SaadMassachusetts Institute of Technology, Alexander K. LewMassachusetts Institute of Technology, USA, Vikash MansinghkaMIT Media Attached | |||||||||||||||||||||||||||||||||||||||||
15:00 - 15:20 Talk | Media Attached |