Michael Doumpos

Multicriteria classification and sorting methods: A literature review

Abstract The assignment of alternatives (observations/objects) into predefined homogenous groups is a problem of major practical and research interest. This type of problem is referred to as classification or sorting, depending on whether the groups are nominal or ordinal. Methodologies for addressing classification and sorting problems have been developed from a variety of research disciplines, including statistics/econometrics, artificial intelligent and operations research. The objective of this paper is to review the research conducted on the framework of the multicriteria decision aiding (MCDA). The review covers different forms of MCDA classification/sorting models, different aspects of the model development process, as well as real-world applications of MCDA classification/sorting techniques and their software implementations.

A Multicriteria Decision Aid Methodology for Sorting Decision Problems: The Case of Financial Distress

Sorting problems constitute a major part of real world decisions, where a set of alternative actions (solutions) must be classified into two or more predefined classes. Multicriteria decision aid (MCDA) provides several methodologies, which are well adapted in sorting problems. A well known approach in MCDA is based on preference disaggregation which has already been used in ranking problems, but it is also applicable in sorting problems. The UTADIS (UTilités Additives DIScriminantes) method, a variant of the UTA method, based on the preference disaggregation approach estimates a set of additive utility functions and utility profiles using linear programming techniques in order to minimize the misclassification error between the predefined classes in sorting problems. This paper presents the application of the UTADIS method in two real world classification problems concerning the field of financial distress. The applications are derived by the studies of Slowinski and Zopounidis (1995), and Dimitras et al. (1999). The obtained results depict the superiority of the UTADIS method over discriminant analysis, and they are also comparable with the results derived by other multicriteria methods.

Detecting Falsified Financial Statements: A Comparative Study using Multicriteria Analysis and Multivariate Statistical Techniques

Falsifying financial statements involves the manipulation of financial accounts by overstating assets, sales and profit, or understating liabilities, expenses or losses. This paper explores the effectiveness of an innovative classification methodology in detecting firms that issue falsified financial statements (FFS) and the identification of the factors associated to FFS. The methodology is based on the concepts of multicriteria decision aid (MCDA) and the application of the UTADIS classification method (UTilités Additives DIScriminantes). A sample of 76 Greek firms (38 with FFS and 38 non-FFS) described over ten financial ratios is used for detecting factors associated with FFS. A jackknife procedure approach is employed for model validation and comparison with multivariate statistical techniques, namely discriminant and logit analysis. The results indicate that the proposed MCDA methodology outperforms traditional statistical techniques which are widely used for FFS detection purposes. Furthermore, the results indicate that the investigation of financial information can be helpful towards the identification of FFS and highlight the importance of financial ratios such as the total debt to total assets ratio, the inventories to sales ratio, the net profit to sales ratio and the sales to total assets ratio.

A multicriteria classification approach based on pairwise comparisons

Abstract Classification refers to the assignment of a set of alternatives into predefined homogenous classes. Most of the existing classification methodologies are based on absolute comparisons among the alternatives and some reference profiles (cut-off points) that discriminate the classes. This paper proposes a new approach that involves pairwise comparisons based on the multicriteria decision aid (MCDA) paradigm. The basis of the methodology is a preference relation that is used to perform pairwise comparisons among the alternatives. The criteria weights used to construct the preference relation are specified using a reference set of alternatives (training sample) on the basis of linear programming techniques.

Credit risk assessment using a multicriteria hierarchical discrimination approach: A comparative analysis

Corporate credit risk assessment decisions involve two major issues: the determination of the probability of default and the estimation of potential future benefits and losses for credit granting. The former issue is addressed by classifying the firms seeking credit into homogeneous groups representing different levels of credit risk. Classification/discrimination procedures commonly employed for such purposes include statistical and econometric techniques. This paper explores the performance of the M.H.DIS method (Multi-group Hierarchical DIScrimination), an alternative approach that originates from multicriteria decision aid (MCDA). The method is used to develop a credit risk assessment model using a large sample of firms derived from the loan portfolio of a leading Greek commercial bank. A total of 1411 firms are considered in both training and holdout samples using financial information through the period 1994–1997. A comparison with discriminant analysis (DA), logit analysis (LA) and probit analysis (PA) is also conducted to investigate the relative performance of the M.H.DIS method as opposed to traditional tools used for credit risk assessment. 2002 Elsevier Science B.V. All rights reserved.

PREFDIS: a multicriteria decision support system for sorting decision problems

Abstract This paper, following the methodological framework of multicriteria decision aid (MCDA), presents the PREFDIS (PREFerence DIScrimination) multicriteria decision support system to study sorting decision problems. The main characteristic and a major advantage of the system is the incorporation into its model base of four MCDA methods originating from the preference-disaggregation approach, namely the UTADIS method (UTilites Additives DIScriminantes) and three of its variants, referred to as UTADIS I, UTADIS II and UTADIS III. Using these methods, the decision maker (DM) can develop interactively powerful additive utility models to sort a set of alternatives into two or more predefined classes as accurately as possible, based on different sorting techniques. Furthermore, the system provides enriched preference modeling capabilities, including the modeling of non-monotone preferences. The friendly window-based user interface of the system enables the decision maker/user to take full advantage of the capabilities of the system in order to make effective real-time decisions. Scope and purpose The sorting problem refers to the assignment of a finite set of alternatives (actions, objects) to predefined ordered classes. Several real-world decision problems are addressed through the sorting approach, including financial decision-making problems, environmental decisions, marketing decisions, and even medical decisions (medical diagnosis). For several decades the sorting (discrimination) among two or more sets of objects has been studied from the multivariate statistical point of view. Recently, the possibilities of new approaches such as expert systems, neural networks, mathematical programming, multicriteria decision aid (MCDA), etc., have been explored, in order to study the sorting problem within a more flexible framework and to develop sorting models with higher discriminating and predicting ability. This paper presents the PREFDIS (PREFerence DISiscrimination) multicriteria decision support system for the study of sorting decision problems. The system incorporating four MCDA sorting methods enables the decision maker to develop interactively, in real time, additive utility models to sort a set of alternatives into two or more predefined classes.

An evolutionary approach to construction of outranking models for multicriteria classification: The case of the ELECTRE TRI method

Outranking methods constitute an important class of multicriteria classification models. Often, however, their implementation is cumbersome, due to the large number of parameters that the decision maker must specify. Past studies tried to address this issue using linear and nonlinear programming, to elicit the necessary preferential information from assignment examples. In this study, an evolutionary approach, based on the differential evolution algorithm, is proposed in the context of the ELECTRE TRI method. Computational results are given to test the effectiveness of the methodology and the quality of the obtained models.

Bankruptcy prediction with neural logic networks by means of grammar-guided genetic programming

The paper demonstrates the efficient use of hybrid intelligent systems for solving the classification problem of bankruptcy. The aim of the study is to obtain classification schemes able to predict business failure. Previous attempts to form efficient classifiers for the same problem using intelligent or statistical techniques are discussed throughout the paper. The application of neural logic networks by means of genetic programming is proposed. This is an advantageous approach enabling the interpretation of the network structure through set of expert rules, which is a desirable feature for field experts. These evolutionary neural logic networks are consisted of an innovative hybrid intelligent methodology, by which evolutionary programming techniques are used for obtaining the best possible topology of a neural logic network. The genetic programming process is guided using a context-free grammar and indirect encoding of the neural logic networks into the genetic programming individuals. Indicative classification results are presented and discussed in detail in terms of both, classification accuracy and solution interpretability.

Assessing profitability factors in the Greek banking system: A multicriteria methodology

The increasing competition in the national and international banking markets, the changeover towards monetary union and the new technological innovations herald major changes in the banking environment, and challenge all banks to make timely preparations in order to enter into the new competitive monetary and financial environment. Therefore, it is interesting to investigate the effectiveness of Greek banks, as it is valued by the financial markets, i.e. the greater the efficacy the higher the competitiveness and vice versa. Taking into consideration the bank assets, we distinguish banks into small and large ones. Finding factors that make the differences in such effectiveness may explain the effective advantage of these two types of financial institutions and help us understand the ‘financial intermediation’ industry in Greece better. Based on their size, a classification of Greek banks, in a multivariate environment, according to the return and operation factors for the years 1990‐1999 takes place. In order to investigate the differences of profitability and efficiency between small and large Greek banks, as well as the factors of profitability and operation related with the size of banks, a multicriteria methodology has been used. The results of this paper may help us determine the key success (or failure) factors of these two categories of Greek banks as well as the responsible banking decision-makers for future readjustments.

Knowledge acquisition and representation for expert systems in the field of financial analysis

Abstract Knowledge acquisition and representation has been characterised as the major bottleneck in the development of expert systems (Barr & Geigenbaum, 1982), especially in problem domains of high complexity. Financial analysis is one of the most complicated practical problems, where the expert systems technology is highly applicable, mainly because of its symbolic reasoning and its explanation capabilities. The aim of this paper is to present a complete methodology for knowledge acquisition and representation for expert systems development in the field of financial analysis. This methodology has been implemented in the development of the FINEVA multicriteria knowledge-based decision support system for the assessment of corporate performance and viability. The application of this methodology in the development of the FINEVA system is presented.