MCDM'11 - paper no. 15
THE DOMINANCE-BASED ROUGH SET APPROACH (DRSA) APPLIED TO BANKRUPCY PREDICTION MODELING FOR SMALL AND MEDIUM BUSINESSES
Kazimierz Zaraś
Abstract:
The preferential information given in the form of ranking or classification examples is more natural than those given in the form of functional parameters or the relational model of the preferences. Nevertheless, processing of these data cause certain difficulties related to a lack of coherence and contradictions in these examples. These contradictions often result from granularity of description language, inaccuracy or uncertainty of the information which makes the decision maker hesitate before the decision making. The model of the preferences will not correct or ignore these contradictions, but rather consider them to release a certain doubtful part of them. Then, exploitation of this model within the framework of decisional problems will lead to unquestionable and possible recommendations. The Rough Set Theory takes into account this postulate making the contradiction analyze possible. This theory was introduced in the early 1980s by Polish researcher Z. Pawlak and developed by S. Greco, B. Matarazzo and R. Slowinski as the Data-based Rough Set Approach (DRSA). In this proposal we will apply the DRSA to hybrid bankruptcy prediction modeling for small businesses. In this modeling the discrimination analysis results are used to explain the decision rules obtained from regional experts.
Keywords:
Multi Criteria Decision Analysis (MCDA), Preference Modelling, Discriminate Analysis, Hybrid Model, Rough Set Theory, Dominance-Based Rough Set Approach (DRSA).
Reference index:
Kazimierz Zaraś, (2011), THE DOMINANCE-BASED ROUGH SET APPROACH (DRSA) APPLIED TO BANKRUPCY PREDICTION MODELING FOR SMALL AND MEDIUM BUSINESSES , Multiple Criteria Decision Making (6), pp. 287-295
download