RUS  ENG
Full version
JOURNALS // Artificial Intelligence and Decision Making // Archive

Artificial Intelligence and Decision Making, 2019 Issue 4, Pages 47–60 (Mi iipr188)

This article is cited in 1 paper

Decision analysis

General approach to multiattribute classification problems based on the verbal decision analysis

E. M. Furems

Federal Research Center "Computer Science and Control" of Russian Academy of Sciences, Moscow, Russia

Abstract: There are a number of methods for ordinal, nominal-ordinal and nominal classification problems within the Verbal Decision Analysis (VDA) paradigm. The VDA-based general approach to all such types of classification problems is proposed. This approach is implemented in the STEPCLASS method, which retains all advantages of above mentioned methods and eliminates their drawbacks. It includes, without limitations, such techniques as (1) initial domain structuring and defining the type of a classification problem; (2) flexible strategy of a dialogue with an expert in order to reveal the complete and consistent set of his/her classification rules with possibility of the problem domain structure extension; and (3) decomposition of large-size classification problem with subsequent aggregation of classification rules over the set of all attributes.

Keywords: verbal decision analysis, multiattribute classification, problem domain structuring, completeness and consistency of classification rules, classification problems with hierarchical structure and/or large-size classification problems.

DOI: 10.14357/20718594190406


 English version:
, 2020, 47:5, 304–313

Bibliographic databases:


© Steklov Math. Inst. of RAS, 2024