SPLASH 2021
Sun 17 - Fri 22 October 2021 Chicago, Illinois, United States
Thu 21 Oct 2021 15:40 - 15:55 at Zurich F - PLDI 2020 Papers 3 Chair(s): Suresh Jagannathan

This paper introduces the MCML approach for empirically studying
the learnability of relational properties that can be
expressed in the well-known software design language Alloy. A
key novelty of MCML is quantification of the performance of and semantic
differences among trained machine learning (ML) models, specifically
decision trees, with respect to entire (bounded) input spaces,
and not just for given training and test
datasets (as is the common practice). MCML reduces the quantification
problems to the classic complexity theory problem of model counting, and employs state-of-the-art model
counters. The results show that relatively simple
ML models can achieve surprisingly high performance (accuracy and F1-score)
when evaluated in the common setting of using
training and test datasets – even when the training dataset is
much smaller than the test dataset – indicating the seeming
simplicity of learning relational properties. However, MCML
metrics based on model counting show that the performance can degrade
substantially when tested against the entire (bounded) input space,
indicating the high complexity of precisely learning these properties,
and the usefulness of model counting in quantifying the true performance.

Thu 21 Oct

Displayed time zone: Central Time (US & Canada) change

15:40 - 17:00
PLDI 2020 Papers 3SIGPLAN Papers at Zurich F
Chair(s): Suresh Jagannathan Purdue University
15:40
15m
Talk
A Study of the Learnability of Relational Properties: Model Counting Meets Machine Learning (MCML)
SIGPLAN Papers
Muhammad Usman University of Texas at Austin, USA, Wenxi Wang University of Texas at Austin, Marko Vasic University of Texas at Austin, USA, Kaiyuan Wang Google, Inc., Haris Vikalo University of Texas at Austin, USA, Sarfraz Khurshid University of Texas at Austin
15:55
15m
Talk
Data-Driven Inference of Representation Invariants
SIGPLAN Papers
Anders Miltner The University of Texas at Austin, Texas, USA, Saswat Padhi Amazon Web Services, USA, Todd Millstein University of California, Los Angeles, David Walker Princeton University, USA
16:10
15m
Talk
Reconciling Enumerative and Deductive Program Synthesis
SIGPLAN Papers
Kangjing Huang Purdue University, USA, Xiaokang Qiu Purdue University, USA, Peiyuan Shen Purdue University, USA, Yanjun Wang Purdue University, USA
16:25
15m
Talk
Synthesizing Structured CAD Models with Equality Saturation and Inverse Transformations
SIGPLAN Papers
Chandrakana Nandi Certora, inc., Max Willsey University of Washington, Adam Anderson University of Washington, James R. Wilcox University of Washington, Eva Darulova Uppsala University, Dan Grossman University of Washington, Zachary Tatlock University of Washington
16:40
20m
Live Q&A
Discussion, Questions and Answers
SIGPLAN Papers