Regularized estimation for preference disaggregation in multiple criteria decision making

Disaggregation methods have been extensively used in multiple criteria decision making to infer preferential information from reference examples, using linear programming techniques. This paper proposes simple extensions of existing formulations, based on the concept of regularization which has been introduced within the context of the statistical learning theory. The properties of the resulting new formulations are analyzed for both ranking and classification problems and experimental results are presented demonstrating the improved performance of the proposed formulations over the ones traditionally used in preference disaggregation analysis.     

Learning by developing knowledge networks

A central challenge for research on how we should prepare students to manage crossing boundaries between different knowledge settings in life long learning processes is to identify those forms of knowledge that are particularly relevant here. In this paper, we develop by philosophical means the concept of adialectical system as a general framework to describe the development of knowledge networks that mark the starting point for learning processes, and we use semiotics to discuss (a) the epistemological thesis that any cognitive access to our world of objects is mediated by signs and (b)diagrammatic reasoning andabduction as those forms of practical knowledge that are crucial for the development of knowledge networks. The richness of this theoretical approach becomes evident by applying it to an example of learning in a biological research context. At the same time, we take a new look at the role of mathematical knowledge in this process.     

New approaches to statistical learning theory

We present new tools from probability theory that can be applied to the analysis of learning algorithms. These tools allow to derive new bounds on the generalization performance of learning algorithms and to propose alternative measures of the complexity of the learning task, which in turn can be used to derive new learning algorithms.     

考虑双参照点和Choquet积分的新型研发机构绩效评估方法及应用

针对新型研发机构绩效评估值不确定、指标间存在关联、投资主体多元且有风险偏好、指标集权重未知等绩效问题,提出基于双参照点和Choquet积分的绩效评估方法。首先,结合新型研发机构的特点,构建绩效评估指标体系;其次,基于累积前景理论设计同行-期望双参照点;然后,构建指标间直接关联矩阵,依据K-可加模糊测度和平均边际贡献Banzhaf值优化求解指标集权重,代入Choquet积分方法和参照点权重公式计算各新型研发机构的综合绩效评估值,从而实现机构排序和问题分析。该方法考虑了新型研发机构的特点、数据不确定性、指标关联性、风险偏好性、权重未知性等因素,更加符合实际情况,案例验证了方法的有效性和实用性。     

Application of multi-attribute utility theory to measure social preferences for health states

A four-attribute health state classification system designed to uniquely categorize the health status of all individuals two years of age and over is presented. A social preference function defined over the health state classification system is required. Standard multi-attribute utility theory is investigated for the task, problems are identified and modifications to the standard method are proposed. The modified methods is field tested in a survey research project involving 112 home interviews. Results are presented and discussed in detail for both the social preference function and the performance of the modified method. A recommended social preference function is presented, complete with a range of uncertainty. The modified method is found to be applicable to the task--no insurmountable difficulties are encountered. Recommendations are presented, based on our experience, for other investigators who may be interested in reapplying the method in other studies.     

Optimization via simulation for logistic systems planning and control

This text summarizes the author’s PhD thesis, presented in February 2007 at the University of Calabria and supervised by Roberto Musmanno and Pasquale Legato. The work deals with the field of optimization via simulation. A special emphasis is put on problems with discrete decision variables because they are especially relevant in engineering applications (e.g., in the design of logistic systems). In particular, the thesis focuses on a critical decision problem at marine container terminals. For tackling the problem, an optimization via simulation procedure is presented. Computational results are obtained by using high performance computing systems. The thesis is written in Italian and is available from the author upon request.      

A simulation optimization method that considers uncertainty and multiple performance measures

Simulation optimization provides a structured approach to system design and configuration when analytical expressions for input/output relationships are unavailable. This research focuses on the development of a new simulation optimization technique applicable to systems having multiple performance measures. The aim of this research is to incorporate a simulation end user’s preference towards risk and uncertainty into the search process for the best decision alternative. Automation of the optimization procedure is a necessity. Therefore, this paper proposes a simulation optimization method that involves a preference model, specifically adapted for decision making with simulation models.     

Was it good for them? Reflections on the use of MENTOR and changes in course design using general performance measures

This paper considers the use of general performance measures in evaluating specific planning and design decisions in higher education and reflects on the students' learning process. Specifically, it concerns the use of the MENTOR multimedia computer aided learning package for helping students learn about OR as part of a general business degree. It includes the transfer of responsibility for a learning module to a new staff member and a change from a single tutor to a system involving multiple tutors. Student satisfaction measures, learning outcome measures and MENTOR usage patterns are examined in monitoring the effects of the changes in course delivery. The results raise some questions about the effectiveness of general performance measures in supporting specific decisions relating to course design and planning.     

Rule learning: Ordinal prediction based on rough sets and soft-computing

This work promotes a novel point of view in rough set applications: rough sets rule learning for ordinal prediction is based on rough graphical representation of the rules. Our approach tackles two barriers of rule learning. Unlike in typical rule learning, we construct ordinal prediction with a mathematical approach, rough sets, rather than purely rule quality measures. This construction results in few but significant rules. Moreover, the rules are given in terms of ordinal predictions rather than as unique values. This study also focuses on advancing rough sets theory in favor of soft-computing. Both theoretical and a designed architecture are presented. The features of our proposed approach are illustrated using an experiment in survival analysis. A case study has been performed on melanoma data. The results demonstrate that this innovative system provides an improvement of rule learning both in computing performance for finding the rules and the usefulness of the derived rules.     

Design of cooperative algorithms for multi-objective optimization: application to the flow-shop scheduling problem

This is a summary of the main results presented in the author’s PhD thesis. This thesis was supervised by El-Ghazali Talbi, and defended on 21 June 2005 at the University of Lille (France). It is written in French and is available at http://www.lifl.fr/~basseur/These.pdf. This work deals with the conception of cooperative methods in order to solve multi-objective combinatorial optimization problems. Many cooperation schemes between exact and/or heuristic methods have been proposed in the literature. We propose a classification of such schemes. We propose a new heuristic called adaptive genetic algorithm (AGA), that is designed for an efficient exploration of the search space. We consider several cooperation schemes between AGA and other methods (exact or heuristic). The performance of these schemes are tested on a bi-objective permutation flow-shop scheduling problem, in order to evaluate the interest of each type of cooperation.     

Non-classical preference models in combinatorial problems: Models and algorithms for graphs

This is a summary of the most important results presented in the authors PhD thesis (Spanjaard 2003). This thesis, written in French, was defended on 16 December 2003 and supervised by Patrice Perny. A copy is available from the author upon request. This thesis deals with the search for preferred solutions in combinatorial optimization problems (and more particularly graph problems). It aims at conciliating preference modelling and algorithmic concerns for decision aiding.Received: March 2004, MSC classification: 91B06, 90C27, 90B40, 16Y60     

A framework for robust and flexible optimisation using metaheuristics

This is a summary of the most important results presented in the authors PhD thesis. This thesis, written in English, was defended on 13 June 2003 and supervised by Johan Springael and Gerrit K. Janssens. A copy is available from the author upon request. This PhD thesis focuses on stochastic problems and develops a framework to find robust and flexible solutions of such problems. The framework is applied to several well-known problems in the realm of supply chain design. Besides this framework, the thesis contains several other important contributions. A new type of robustness and flexibility is proposed, that expresses the need for solutions to remain approximately the same when changes occur in the problem data. Several distance measures are developed to calculate the distance (or similarity) between solutions of different permutation type problems. A new type of genetic algorithm is also proposed, that uses a distance measure to maintain a diverse population of high-quality solutions.Received: June 2003     

A computationally fast variable importance test for random forests for high-dimensional data

Random forests are a commonly used tool for classification and for ranking candidate predictors based on the so-called variable importance measures. These measures attribute scores to the variables reflecting their importance. A drawback of variable importance measures is that there is no natural cutoff that can be used to discriminate between important and non-important variables. Several approaches, for example approaches based on hypothesis testing, were developed for addressing this problem. The existing testing approaches require the repeated computation of random forests. While for low-dimensional settings those approaches might be computationally tractable, for high-dimensional settings typically including thousands of candidate predictors, computing time is enormous. In this article a computationally fast heuristic variable importance test is proposed that is appropriate for high-dimensional data where many variables do not carry any information. The testing approach is based on a modified version of the permutation variable importance, which is inspired by cross-validation procedures. The new approach is tested and compared to the approach of Altmann and colleagues using simulation studies, which are based on real data from high-dimensional binary classification settings. The new approach controls the type I error and has at least comparable power at a substantially smaller computation time in the studies. Thus, it might be used as a computationally fast alternative to existing procedures for high-dimensional data settings where many variables do not carry any information. The new approach is implemented in the R package vita.     

On the possibility theory-based semantics for logics of preference

In this paper, we study the semantics for the logics of preference based on possibility theory. Possibility distributions representing the preference between worlds are associated with the possible world models for dynamic logics. Then the preference between actions are determined by comparing some measures of their consequences. We define different logics of preference by considering the comparisons of possibility measures and guaranteed possibility measures. Some properties of the proposed logics are studied and their relationships with deontic logics are also considered.     

Solving the capacitated location-routing problem

This is a summary of the main results presented in the author’s Ph.D thesis, available at http://prodhonc.free.fr/homepage. This thesis, written in French, was supervised by Christian Prins and Roberto Wolfler-Calvo, and defended on 16 October 2006 at the Université de Technologie de Troyes. Several new approaches are proposed to solve the capacitated location-routing problem (CLRP): heuristic, cooperative and exact methods. Their performances are tested on various kinds of instances with capacitated vehicles and capacitated or uncapacitated depots.      

Matrix representation of a binary relation using fuzzy and artificial learning theory. An algorithm which uses the potential functions learning rule

One of the most significant problems in economic domain is the dispose of human preference and choice forecasting. Recently, the economists have focused their researches to use the fuzzy concepts and the artificial learning procedures in the theory of economic choice. This paper extends the work done in this direction and offers a new algorithm for finding the matrix representation of the fuzzy binary relation which describes a preference relation.     

Distances in evidence theory: Comprehensive survey and generalizations

The purpose of the present work is to survey the dissimilarity measures defined so far in the mathematical framework of evidence theory, and to propose a classification of these measures based on their formal properties. This research is motivated by the fact that while dissimilarity measures have been widely studied and surveyed in the fields of probability theory and fuzzy set theory, no comprehensive survey is yet available for evidence theory. The main results presented herein include a synthesis of the properties of the measures defined so far in the scientific literature; the generalizations proposed naturally lead to additions to the body of the previously known measures, leading to the definition of numerous new measures. Building on this analysis, we have highlighted the fact that Dempster’s conflict cannot be considered as a genuine dissimilarity measure between two belief functions and have proposed an alternative based on a cosine function. Other original results include the justification of the use of two-dimensional indexes as (cosine; distance) couples and a general formulation for this class of new indexes. We base our exposition on a geometrical interpretation of evidence theory and show that most of the dissimilarity measures so far published are based on inner products, in some cases degenerated. Experimental results based on Monte Carlo simulations illustrate interesting relationships between existing measures.     

基于网络混合学习策略的股市投资者行为演化模型及仿真

针对投资者有限理性决策行为的模仿学习等特质,运用复杂网络、贝叶斯学习和社会学习理论,分别对个人投资者和机构投资者的投资行为决策机制进行系统性分析,由此构建基于网络混合学习策略的投资者行为演化机理和数理模型,并进行数理解析和模拟仿真。研究发现,投资者金融关联网络的拓扑结构对个人投资者行为动态演变具有较强影响,但对机构投资者的影响较弱;个人投资者的学习策略偏好影响其行为动态演变均衡状态的稳定性较为显著;机构投资者获取的私有信息精确度与其行为动态演变的均衡状态之间存在显著的非线性关联;投资者的初始行为状态对其投资行为动态演变均衡状态的影响效应均较弱。