Multicriteria classification and sorting methods: A literature review

ELECTRE: A comprehensive literature review on methodologies and applications

Computational Intelligence and Feature Selection provides a high level audience with both the background and fundamental ideas behind feature selection with an emphasis on those techniques based on rough and fuzzy sets, including their hybridizations It introduces set theory, fuzzy set theory, rough set theory, and fuzzy-rough set theory, and illustrates the power and efficacy of the feature selection described through the use of real-world applications and worked examples Program files implementing major algorithms covered, together with the necessary instructions and datasets, are available on the Web

Computational Intelligence and Feature Selection: Rough and Fuzzy Approaches

We present main characteristics of ELECTRE (ELimination Et Choix Traduisant la REalite - ELimination and Choice Expressing the REality) family methods, designed for multiple criteria decision aiding. These methods use as a preference model an outranking relation on the set of actions—it is constructed in result of concordance and nondiscordance tests involving a specific input preference information. After a brief description of the constructivist conception in which the ELECTRE methods are inserted, we present the main features of these methods. We discuss such characteristic features as the possibility of taking into account positive and negative reasons in the modelling of preferences, without requiring commensurable performance scales; the use of discriminating thresholds for taking into account the imperfect knowledge of data; the absence of systematic compensation between ‘gains’ and ‘losses’. The main weaknesses are also presented. Then, some aspects related to new developments are outlined. These are related to some new methodological developments, new procedures, axiomatic analysis, software tools and several other aspects. This paper is an updated version of a chapter published by the authors under the title ‘ Electre Methods: Main Features and Recent Developments’ in C. Zopounidis and P. Pardalos (Editors): Handbook of Multicriteria Analysis, Springer, Berlin 2010, pp. 51–89. Copyright © 2012 John Wiley & Sons, Ltd.

An Overview of ELECTRE Methods and their Recent Extensions

https://www.econstor.eu/bitstream/10419/85773/1/02028.pdf

Meta-analysis of the impact of fiscal policies on long-run growth

This paper briefly describes ROSE software package. It is an interactive, modular system designed for analysis and knowledge discovery based on rough set theory in 32-bit operating systems on PC computers. It implements classical rough set theory as well as its extension based on variable precision model. It includes generation of decision rules for classification systems and knowledge discovery.

ROSE - Software Implementation of the Rough Set Theory

This article briefly describes the process of data exploration based on rough set theory and also proposes ROSE system as a useful toolkit for doing such data analysis on PC computers.

Rough Set Based Data Exploration Using ROSE System

Solving multicriteria decision problems, like choice and ranking, requires the use of decision maker's DM's (Decision Maker's) preference model. In this paper, we investigate some issues of equivalence between the preference model in terms of "if..., then..." decision rules and a concordance-discordance preference model based on the use of the outranking relation. The decision rule model is attractive for at least two reasons: (i) it is intelligible and speaks the language of the DM, (ii) the preference information coming from the DM is a set of decision examples. The decision rules are induced from rough approximations of the preference relation specified in decision examples. Then, from the set of decision rules representing the DM's preferences, criteria weights and veto thresholds are inferred, setting up an equivalent preference model following from concordance-discordance tests proposed in ELECTRE methods. A simple example will illustrate the interest of such an equivalence.

Searching for an equivalence between decision rules and concordance-discordance preference model in multicriteria choice problems

The relationships between the chemical structure and the antielectrostatic effect of 180 imidazolium chlorides were analysed using the method of rough sets. The antielectrostatic activity was determined by measurements of the surface resistance. Using the rough sets approach the smallest set of condition attributes significant for high quality of classification has been found. The resulting decision rules describe relations between the structure and the antielectrostatic properties of imidazolium chlorides in terms of significant condition attributes. This may be helpful in predicting the structures of the new antielectrostatic compounds to be synthesized.

Application of the Rough Set Theory in Structure Activity Relationships of Antielectrostatic Imidazolium Compounds

The research concerns analysis of cause-effect relationships occurring between condition attributes describing structure and surface-active properties and decision attributes (antimicrobial activity) of alkyl(alkylthiomethyl)dimethylammonium chlorides and alkyl(alkoxymethyl)dimethylammonium chlorides. 72 ammonium chlorides were described by 7 condition attributes concerning structure and 7 condition attributes concerning surface-active properties. Antimicrobial activity was expressed by minimal inhibitory concentration (MIC) against strains of Pseudomonas aeruginosa ATCC 15442 (classification I) and Candida albicans ATCC 10231 (classification II). The ROSE 2 (Rough Sets Data Explorer) and LEM 2 (Minimal Set Covering Algorithm) software were used to perform the analysis. Attributes with continuous values were discretized. 45 decision rules were generated in the classification I, and 38 rules in the classification II. Those described the most important dependencies between structure, surface-active properties of analyzed chlorides and their antimicrobial activity. Strong decision rules supported by large number of objects provide guidelines which may facilitate synthesis of new compounds with preferable antimicrobial properties.

Bartlomiej Predki

Papers

ROSE - Software Implementation of the Rough Set Theory

Rough Set Based Data Exploration Using ROSE System

Searching for an equivalence between decision rules and concordance-discordance preference model in multicriteria choice problems

Application of the Rough Set Theory in Structure Activity Relationships of Antielectrostatic Imidazolium Compounds

Analysis of Relationships Between Structure, Surface Properties, and Antimicrobial Activity of Quaternary Ammonium Chlorides