On the KL divergence of probability mixtures for belief contraction

Kinzang Chhogyal*, Abhaya Nayak, Abdul Sattar

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contributionpeer-review

1 Citation (Scopus)


Probabilistic belief change is an operation that takes a probability distribution representing a belief state along with an input sentence representing some information to be accommodated or removed, and maps it to a new probability distribution. In order to choose from many such mappings possible, techniques from information theory such as the principle of minimum cross-entropy have previously been used. Central to this principle is the Kullback-Leibler (KL) divergence. In this short study, we focus on the contraction of a belief state P by a belief a, which is the process of turning the belief a into a non-belief. The contracted belief state Pa can be represented as a mixture of two states: the original belief state P, and the resultant state P*¬a of revising P by ¬a. Crucial to this mixture is the mixing factor ∊ which determines the proportion of P and P*¬a that are to be used in this process. We show that once ∊ is determined, the KL divergence of Pa from P is given by a function whose only argument is ∊. We suggest that ∊ is not only a mixing factor but also captures relevant aspects of P and P*¬a required for computing the KL divergence.

Original languageEnglish
Title of host publicationKI 2015: Advances in Artificial Intelligence
Subtitle of host publication38th Annual German Conference on AI, Dresden, Germany, September 21-25, 2015, Proceedings
EditorsSteffen Hölldobler, Markus Krötzsch, Rafael Peñaloza, Sebastian Rudolph
Place of PublicationCham
PublisherSpringer, Springer Nature
Number of pages7
ISBN (Electronic)9783319244891
ISBN (Print)9783319244884
Publication statusPublished - 2015
Event38th Annual German Conference on Advances in Artificial Intelligence, AI 2015 - Dresden, Germany
Duration: 21 Sept 201525 Sept 2015

Publication series

NameLecture Notes in Computer Science
PublisherSpringer International Publishing
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Other38th Annual German Conference on Advances in Artificial Intelligence, AI 2015


Dive into the research topics of 'On the KL divergence of probability mixtures for belief contraction'. Together they form a unique fingerprint.

Cite this