Probabilistic generalization of formal concepts

An inductive probabilistic approach to formal concept analysis (FCA) is proposed in which probability on formal contexts is considered; probabilistic formal concepts that have predictive force are defined: nonclassified objects can be assigned to earlier found probabilistic formal concepts; random a...

Full description

Saved in:
Bibliographic Details
Published in:Programming and computer software 2012-09, Vol.38 (5), p.219-230
Main Authors: Vityaev, E. E., Demin, A. V., Ponomaryov, D. K.
Format: Article
Language:English
Subjects:
Artificial Intelligence
Computer Science
Operating Systems
Random noise
Software Engineering
Software Engineering/Programming and Operating Systems
Statistical analysis
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:An inductive probabilistic approach to formal concept analysis (FCA) is proposed in which probability on formal contexts is considered; probabilistic formal concepts that have predictive force are defined: nonclassified objects can be assigned to earlier found probabilistic formal concepts; random attributes are eliminated from probabilistic formal concepts; probabilistic formal concepts are robust with respect to data noise. A result of experiment is presented in which formal concepts (in their standard definition in FCA) are first distorted by random noise and then recovered by detecting probabilistic formal concepts.
ISSN:0361-7688
1608-3261
DOI:10.1134/S0361768812050076