|
Activity Number:
|
217
- Contributed Poster Presentations: Section on Statistical Computing & Statistics in Sports
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2020 : 10:00 AM to 2:00 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
|
Abstract #312569
|
|
|
Title:
|
Practical Approaches to Accelerating Exact Conditional Inference in Discrete Exponential Families
|
|
Author(s):
|
Grant Innerst*
|
|
Companies:
|
Shippensburg University
|
|
Keywords:
|
Algebraic Statistics
|
|
Abstract:
|
Conducting exact conditional inference in discrete exponential family mod- els can be challenging because the distribution of interest typically cannot be normalized. In theory, the use of a Markov chain Monte Carlo (MCMC) algorithm makes it possible to compute probabilities such as p-values. In practice, the performance of the MCMC algorithm can itself be prohibitive and depends heavily on the variant of MCMC used and details associated with the implementation of the algorithm. In this work we present novel C++ implementations of different strategies to improve its behavior. We also explore extensions to exact unconditional inference. These implementations are available in the open-source R package algstat. This material is based upon work supported by the National Science Foundation under Grant No. 1622449.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2020 program
|
