Advertisement
Published in:
2017 | OriginalPaper | Chapter
Application of Risk Theory Approach to Fuzzy Abduction
Author : V. N. Tsypyschev
Published in: Cybernetics and Mathematics Applications in Intelligent Systems
Publisher: Springer International Publishing
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by (Link opens in a new window)
Abstract
In this article, learning under the absence or incompleteness of some facts or premises about the problem domain is considered. This task does not fall under semi-supervised learning in the classical sense, because there it is assumed that the target signals are known and correct. The assumption of incompleteness is, however, natural for pattern recognition, e.g. for medical diagnostics.
In such a situation, it is natural to base a learning process on abductive reasoning instead of induction or transduction. It is then important to have a quality criterion for the state of knowledge on the object to be studied.
Previously, to reconstruct missing training data, a fuzzy logical approach to the application of the abductive reasoning method was studied. Now, fuzzy abduction is considered from a risk-theoretical point of view.
As a result, in addition to the fuzzy abduction method, a general algorithm is suggested for finding the true state of the object to be studied in the case when known hypotheses about its state are mutually far from each other.
