English
 
User Manual Privacy Policy Disclaimer Contact us
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT
 
 
DownloadE-Mail
  Testing Conditional Independence on Discrete Data using Stochastic Complexity

Marx, A., & Vreeken, J. (2019). Testing Conditional Independence on Discrete Data using Stochastic Complexity. Retrieved from http://arxiv.org/abs/1903.04829.

Item is

Basic

show hide
Item Permalink: http://hdl.handle.net/21.11116/0000-0004-027A-1 Version Permalink: http://hdl.handle.net/21.11116/0000-0004-027B-0
Genre: Paper

Files

show Files
hide Files
:
arXiv:1903.04829.pdf (Preprint), 923KB
Name:
arXiv:1903.04829.pdf
Description:
File downloaded from arXiv at 2019-07-10 09:24 accepted at AISTATS'19, the proposed test was released in the R package SCCI
Visibility:
Public
MIME-Type / Checksum:
application/pdf / [MD5]
Technical Metadata:
Copyright Date:
-
Copyright Info:
-

Locators

show

Creators

show
hide
 Creators:
Marx, Alexander1, Author              
Vreeken, Jilles1, Author              
Affiliations:
1Databases and Information Systems, MPI for Informatics, Max Planck Society, ou_24018              

Content

show
hide
Free keywords: Statistics, Machine Learning, stat.ML,Computer Science, Learning, cs.LG
 Abstract: Testing for conditional independence is a core aspect of constraint-based causal discovery. Although commonly used tests are perfect in theory, they often fail to reject independence in practice, especially when conditioning on multiple variables. We focus on discrete data and propose a new test based on the notion of algorithmic independence that we instantiate using stochastic complexity. Amongst others, we show that our proposed test, SCI, is an asymptotically unbiased as well as $L_2$ consistent estimator for conditional mutual information (CMI). Further, we show that SCI can be reformulated to find a sensible threshold for CMI that works well on limited samples. Empirical evaluation shows that SCI has a lower type II error than commonly used tests. As a result, we obtain a higher recall when we use SCI in causal discovery algorithms, without compromising the precision.

Details

show
hide
Language(s): eng - English
 Dates: 2019-03-122019
 Publication Status: Published online
 Pages: 18 p.
 Publishing info: -
 Table of Contents: -
 Rev. Type: -
 Identifiers: arXiv: 1903.04829
URI: http://arxiv.org/abs/1903.04829
BibTex Citekey: Marx_arXiv1903.04829
 Degree: -

Event

show

Legal Case

show

Project information

show

Source

show