Browsing by Author "Mousseau, Vincent"

Now showing 1 - 8 of 8

A Decomposition Based Minimax Regret Approach for Inverse Multiple Criteria Sorting Problem
(Springer Heidelberg, 2023) Özpeynirci, Özgür; Özpeynirci, Selin; Mousseau, Vincent
Multiple criteria sorting problem aims to assign objects evaluated on multiple criteria to ordered classes. In inverse multiple criteria sorting problem, the class assignments of objects are known and the decision maker can manipulate the scores of objects on criteria by implementing actions. Selected actions enable the improvement of objects' final classification. As the decision maker chooses to implement more actions, better classifications may be obtained. The contribution of this paper is under two-folds. First, we decompose inverse multiple criteria sorting problem into two phases, where phase one is a pre-process that computes the minimum cost required for each feasible object-class pair considering the underlying sorting model. Phase two interacts with the decision maker to analyze the classification and budget related trade-offs, through an assignment model generated with the outputs of phase one. The second contribution is using a modified version of a regret-based approach available in the literature. This modification includes a tighter formulation of the regret model, and an interactive solution approach using a mixed integer program for computing the minimax regret value rather than a branch-and-bound approach. We present an example instance to illustrate the developed ideas and conduct computational tests on randomly generated instances. The simultaneous use of the decomposition approach, tighter formulation and the interactive algorithm reduces the computation time significantly.
Citation - WoS: 17
Citation - Scopus: 19
An Interactive Algorithm for Multiple Criteria Constrained Sorting Problem
(Springer, 2018) Özpeynirci, Selin; Özpeynirci, Özgür; Mousseau, Vincent
In this study, we consider the multiple criteria constrained sorting problem and propose an interactive algorithm. The problem is to assign alternatives evaluated on multiple criteria to sorted categories where categories may have size restrictions. The proposed interactive algorithm determines the eligible categories for each alternative and the alternatives with a single eligible category are assigned to the corresponding categories. In case of no such alternative exists, the algorithm asks the DM to assign an alternative to a category and proceeds. The algorithm has the capability to detect and resolve any inconsistencies that may arise during the iterations. We implement the algorithm on two different real life problems considering two underlying sorting methods, MR-Sort and UTADIS and present the computational results.
Citation - WoS: 6
Citation - Scopus: 6
An Interactive Algorithm for Resource Allocation With Balance Concerns
(Springer, 2021) Özpeynirci, Selin; Özpeynirci, Özgür; Mousseau, Vincent
We consider a resource allocation problem where a decision maker (DM) is to distribute a certain budget among alternative projects in order to create the best portfolio. The DM aims to generate a portfolio with (1) a high return and (2) a balanced distribution of resources among categories. We assume that the DM has a quasiconcave value function and provides no explicit value for the target distribution of resources over categories. We develop an interactive approach that requires the DM to make pairwise comparisons among alternative portfolios. We present the developed approach on an illustrative example and conduct an extensive computational experiment. In majority of the instances, the incumbent solution proposed by the algorithm is either equal or very close to the best solution.
Citation - WoS: 5
Citation - Scopus: 6
An Interactive Approach for Inverse Multiple Criteria Sorting Problem
(Wiley, 2021) Özpeynirci, Özgür; Özpeynirci, Selin; Mousseau, Vincent
Multiple criteria sorting is the problem of assigning objects that are evaluated on multiple criteria to ordered classes. In inverse multiple criteria sorting problem, the assignment of objects to classes is known. The decision maker (DM) can implement actions to improve the scores of objects on criteria, thus enabling the assignment of objects to better classes. In inverse multiple criteria sorting problems, two types of questions arise: (1) What is the set of actions that will result in the desired classification with minimum cost? and 2) What is the set of actions to obtain the best classification without exceeding a predetermined budget? There exist two versions for each question: (i) Simple version, where the parameters of the sorting method are known and (ii) Robust version, where the parameters are not fully known. In this article, we focus on the case where the DM does not provide any budget limitation or desired classifications a priori. By implementing more actions, that is, by spending more money, better classifications may be obtained. We aim to analyse the trade-off between cost and classification, and determine the set of actions that will result in the most preferred cost-classification compromise by the DM. We develop an approach that proceeds interactively with the DM, and determines such most preferred set of actions. The developed approach assumes the simple version and uses MR-Sort as sorting model. We provide a detailed illustrative example on a real life problem and present the progress of the interactive algorithm.
Citation - WoS: 16
Citation - Scopus: 18
Inverse Multiple Criteria Sorting Problem
(Springer, 2018) Mousseau, Vincent; Özpeynirci, Özgür; Özpeynirci, Selin
Multiple criteria sorting problem is to assign objects evaluated with multiple criteria to one of the predefined ordered classes. In this study, we consider the inverse multiple criteria sorting problem (IMCSP), in which it is possible to perform actions which have an impact of objects evaluations, hence on the objects classification. IMCSP aims at determining which action(s) to implement so as to provide guaranties on objects classification. Each action has a corresponding cost and impact on the evaluations of objects on each criterion. In this paper we study IMCSP for three different sorting methods: linear, UTADIS and MR-Sort. We consider two levels of information; (i) the sorting method parameters are known explicitly (simple version), and (ii) assignment examples restrict the set of compatible parameters (robust version). We study two types of problems; first, finding the least costly set of actions that guarantees the objects assignment to desired classes, and second, improving the assignment of objects under a limited budget. For each case, we develop a resolution method based on mathematical programming models. Extensive computational experiments on randomly generated instances show the performance and applicability of the approach.
Citation - WoS: 12
Multi-Criteria Sorting With Category Size Restrictions
(World Scientific Publ Co Pte Ltd, 2017) Koksalan, Murat; Mousseau, Vincent; Özpeynirci, Selin
We consider the multi-criteria sorting problem where alternatives that are evaluated on multiple criteria are assigned into ordered categories. We focus on the sorting problem with category size restrictions, where the decision maker (DM) may have some concerns or constraints on the number of alternatives that should be assigned to some of the categories. We develop an approach based on the UTADIS method that fits an additive utility function to represent the decision maker's preferences. We introduce additional variables and constraints to enforce the restrictions on the sizes of categories. The new formulation reduces the number of binary variables and hence decreases the computational effort compared to the existing approaches in the literature. We further improve the computational efficiency by developing lower and upper bounds on the rank of each alternative in order to narrow down the set of categories that each alternative can be assigned to. We demonstrate our approach on two applications from practice.
Citation - WoS: 33
Citation - Scopus: 40
A New Outranking-Based Approach for Assigning Alternatives To Ordered Classes
(John Wiley & Sons Inc, 2009) Koksalan, Murat; Mousseau, Vincent; Özpeynirci, Özgür; Özpeynirci, Selin
We consider the problem of assigning alternatives evaluated on several criteria into ordered categories C(1), C(2), ..., C(p). This problem is known as the multi-criteria sorting problem and arises in many situations such as classifying countries into different risk levels based on economical and socio-political criteria, evaluating credit applications of bank customers. We are interested in sorting methods that are grounded on the construction Of Outranking relations. Among these, the Electre Tri method requires defining multidimensional profiles that represent the frontier separating consecutive categories C(h) and C(h+1). and assigns an alternative to categories according to how it compares to each of the profiles. The explicit specification of the profiles of consecutive categories can be difficult for decision makers. We develop a new Outranking based sorting method that does not require the explicit definition of profiles. We instead require the decision maker to assign a Subset of reference alternatives to the categories. To assign the remaining alternatives. each such alternative is compared to reference alternatives, and assigned to categories accordingly. (C) 2008 Wiley Periodicals, Inc. Naval Research Logistics 56: 74-85, 2009
Citation - WoS: 3
Citation - Scopus: 3
Portfolio Decision Analysis With a Generalized Balance Approach
(Pergamon-Elsevier Science Ltd, 2022) Özpeynirci, Selin; Özpeynirci, Özgür; Mousseau, Vincent
We consider a resource allocation problem in which the decision maker is responsible for selecting a group of projects to generate a portfolio, with the aim of maximizing the total benefit, and distributing the budget in a balanced way across different categories. We assume that the decision maker defines an interval for the desired share of each category. We propose augmented epsilon-constraint method to generate the nondominated solutions considering two objective functions: total benefit and balance among categories. The solution time increases with the problem size, and augmented epsilon-constraint method fails to find all nondominated solutions. We therefore develop a variable neighborhood search algorithm for the purpose of generating the nondominated frontier. Our computational experiments show that the nondominated frontier can be accurately estimated with respect to different performance measures.