

Link to the official university repository Link to the official HEC repository Michaël Schyns: research
Papers, Working Papers, ThesisSpecial issue: Twelfth Workshop on Models and Algorithms for Planning and Scheduling Problems (MAPSP 2015) Crama, Yves; Goossens, Dries; Leus, Roel; Schyns, Michael; Spieksma, Frits in Journal of Scheduling (2017), 20(6), 543711 This special issue of the Journal of Scheduling contains ten papers presented at the Twelfth Workshop on Models and Algorithms for Planning and Scheduling Problems (MAPSP 2015), held from June 8 to June 12, 2015, in La RocheenArdenne, Belgium. Early Detection of University Students with Potential Difficulties Hoffait, AnneSophie; Schyns, Michael in Decision Support Systems (2017), 101 Using data mining methods, this paper presents a new means of identifying freshmen's profiles likely to face major difficulties to complete their first academic year. Academic failure is a relevant issue at a time when postsecondary education is ever more critical to economic success. We aim at early detection of potential failure using student data available at registration, i.e. school records and environmental factors, with a view to timely and efficient remediation and/or study reorientation. We adapt three data mining methods, namely random forest, logistic regression and artificial neural network algorithms. We design algorithms to increase the accuracy of the prediction when some classes are of major interest. These algorithms are context independent and can be used in different fields. Real data pertaining to undergraduates at the University of Liège (Belgium), illustrates our methodology. MIPbased constructive heuristics for the threedimensional Bin Packing Problem with transportation constraints Paquay, Célia; Limbourg, Sabine; Schyns, Michael; Oliveira, José Fernando in International Journal of Production Research (2017) This article is about seeking a good feasible solution in a reasonable amount of computation time to the threedimensional Multiple Bin Size Bin Packing Problem (MBSBPP). The MBSBPP studied considers additional constraints encountered in real world air transportation situations, such as cargo stability and the particular shape of containers. This MBSBPP has already been formulated as a Mixed Integer linear Programming problem, but as yet only poor results have been achieved for even fairly small problem sizes. The goal of the work this paper describes is to develop heuristics that are able to quickly provide good initial feasible solutions for the MBSBPP. Three methodologies are considered, which are based on the decomposition of the original problem into easier subproblems: the matheuristics RelaxandFix, InsertandFix and Fractional RelaxandFix. They have been parametrised on real data sets and then compared to each other. In particular, two of these techniques show promising results in reasonable computational times. Accounting for Price Endogeneity in Airline Itinerary Choice Models: An Application to Continental U.S. Markets Lurkin, Virginie; Garrow, Laurie; Higgins, Matthew; Newman, Jeffrey; Schyns, Michael in Transportation Research. Part A : Policy & Practice (2017), 100 Network planning models, which forecast the profitability of airline schedules, support many critical decisions, including equipment purchase decisions. Network planning models include an itinerary choice model which is used to allocate air total demand in a city pair to different itineraries. Multinomial logit (MNL) models are commonly used in practice and capture how individuals make tradeoffs among different itinerary attributes; however, none that we are aware of account for price endogeneity. This study formulates an itinerary choice model that is consistent with those used by industry and corrects for price endogeneity using a control function that uses several types of instrumental variables. We estimate our models using database of more than 3 million tickets provided by the Airlines Reporting Corporation. Results based on Continental U.S. markets for May 2013 departures show that models that fail to account for price endogeneity overestimate customers’ value of time and result in biased price estimates and incorrect pricing recommendations. The size and comprehensiveness of our database allows us to estimate highly refined departure time of day preference curves that account for distance, direction of travel, the number of time zones traversed, departure day of week and itinerary type (outbound, inbound or oneway). These time of day preference curves can be used by airlines, researchers, and government organizations in the evaluation of different policies such as congestion pricing. A tailored twophase constructive heuristic for the threedimensional Multiple Bin Size Bin Packing Problem with transportation constraints Paquay, Célia; Limbourg, Sabine; Schyns, Michael in European Journal of Operational Research (2017) This paper considers the threedimensional Multiple Bin Size Bin Packing Problem which consists in packing a set of cuboid boxes into containers of various shapes with minimising unused space. The problem is extended to air cargo where bins are Unit Load Devices, especially designed for fitting in aircraft. We developed a fast constructive heuristic able to manage the different constraints met in transportation. The heuristic is split into two distinct phases. The first phase deals with the packing of boxes into identical bins using an extension of the Extreme Points. During this phase, the fragility, stability and orientations of the boxes are taken into account as well as the special shape of the bins and their weight capacity. The second phase takes into account the multiple types of available bins. If necessary, the best found loading pattern is finally enhanced with respect to weight distribution in a post processing. After parametrisation, computational experiments have been performed on data sets especially designed for this application. The heuristic requires really short computational times to achieve promising results. Early Detection of University Students in Potential Difficulty Hoffait, AnneSophie; Schyns, Michael Eprint/Working paper (2016) Using data mining methods, this paper presents a new means of identifying freshmen's profiles likely to face major difficulties to complete their first academic year. Academic failure is a relevant issue at a time when postsecondary education is ever more critical to economic success. We aim at early detection of potential failure using student data available at registration, i.e. school records and environmental factors, with a view to timely and efficient remediation and/or study reorientation. For the sake of accuracy, we adapt three data mining methods, namely random forest, logistic regression and artificial neural network algorithms. Real data pertaining to undergraduates at the University of Liège (Belgium), illustrates our methodology. A Comparison of Departure Time of Day Formulations Lurkin, Virginie; Garrow, Laurie; Higgins, Matthew; Newman, Jeffrey; Schyns, Michael Eprint/Working paper (2016) Airline passengers’ itinerary choices are influenced by many factors including carriers, prices, the number of connections, and departure times. This paper compares three different methods that have been used to model departure time of day preferences. The first is a discrete formulation that uses indicator variables to represent the hour of departure. The next two methods are based on a continuous formulation that uses a series of sine and cosine functions. One assumes departure time preferences over a 24hour cycle and the other uses shorter cycle lengths that account for fewer departures during certain hours of the day. We compare models using itineraries in the Continental U.S. that are separated by two time zones. Although the discrete formulation fits the data better, the two continuous time of day formulations are preferred as they provide more intuitive predictions and require fewer parameters. Results between the two continuous time of day formulations are similar but differ in how strongly they weight itineraries that depart very early or very late in the day. Based on empirical results, we recommend testing both 24hour and less than 24hour cycle lengths for a particular dataset. Continuous Departure Time of Day Preferences for Continental U.S. Airline Markets Segmented by Distance, Direction of Travel, Number of Time Zones, Day of Week and Itinerary Type Lurkin, Virginie; Garrow, Laurie; Higgins, Matthew; Newman, Jeffrey; Schyns, Michael Eprint/Working paper (2016) Airlines use itinerary choice models to allocate the total number of passengers in a city pair to specific itineraries. In a related paper, we estimated a multinomial logit (MNL) itinerary choice model using database of more than 3 million tickets for Continental U.S. markets provided by the Airlines Reporting Corporation that accounted for price endogeneity. The size and comprehensiveness of our database allowed us to estimate highly refined continuous departure time of day preference curves that account for distance, direction of travel, the number of time zones traversed, departure day of week and itinerary type (outbound, inbound or oneway). This paper and accompanying Excel spreadsheet located at http://garrowlab.ce.gatech.edu contain the results of all model coefficients (including the 1260 time of day parameters) and summarize results in a series of ten figures. These highlyrefined time of day preference curves can be used by airlines, researchers, and government organizations in the evaluation of demandmanagement and other policies. A Mixed Integer Programming formulation for the three dimensional bin packing problem deriving from an air cargo application Paquay, Célia; Schyns, Michael; Limbourg, Sabine in International Transactions in Operational Research (2016), 23(12), 187213 The present paper looks into the problem of optimising the loading of boxes into containers. The goal is to minimise the unused volume. This type of problem belongs to the family of Multiple Bin Size Bin Packing Problems. The approach includes an extensive set of constraints encountered in realworld applications in the threedimensional case: the stability, the fragility of the items, the weight distribution and the possibility to rotate the boxes. It also includes the specific situation in which containers are truncated parallelepipeds. This is typical in the field of air transportation. While most papers on cutting and packing problems describe adhoc procedures, this paper proposes a mixed integer linear program. The validity of this model is tested on small instances. An Ant Colony System for Responsive Dynamic Vehicle Routing Schyns, Michael in European Journal of Operational Research (2015), 245(3), 704718 We present an algorithm based on an Ant Colony System to deal with a broad range of Dynamic Capacitated Vehicle Routing Problems with Time Windows, (partial) Split Delivery and Heterogeneous fleets (DVRPTWSD). Besides the traditional distance criterion, we address the important case of responsiveness. Responsiveness is defined here as completing a delivery as soon as possible, within the time window, such that the client or the truck may restart its activities. This is crucial for many production or service activities in different fields: express parcel deliveries, taxi services, Just in Time production, express repair services, medical care, petrol station replenishment, etc. We develop an interactive webbased solution to allow dispatchers to take new information into account in realtime. The algorithm and its parametrization were tested on real and artificial instances. We first illustrate our approach with a problem submitted by Liege Airport, the 8th biggest cargo airport in Europe. The goal is to develop a decision system to optimize the journey of the refueling trucks. We then consider some classical VRP benchmarks with extensions for more complex problems. The Airline Container Loading Problem with Pickup and Delivery Lurkin, Virginie; Schyns, Michael in European Journal of Operational Research (2015), 244(3), 955965 This paper considers the loading optimization problem for a set of containers and pallets transported into a cargo aircraft that serves multiple airports. Because of pickup and delivery operations that occur at intermediate airports, this problem is simultaneously a Weight & Balance Problem and a Sequencing Problem. Our objective is to minimize fuel and handling operation costs. This problem is shown to be NPhard. We resort to a mixed integer linear program. Based on realworld data from a professional partner (TNT Airways), we perform numerical experiments using a standard B&C library. This approach yields better solutions than traditional manual planning, which results in substantial cost savings. How to optimally load a set of containers into an aircraft Schyns, Michael; Limbourg, Sabine; Laporte, Gilbert Eprint/Working paper (2012) Automatic Aircraft Load Planning Limbourg, Sabine; Schyns, Michael; Laporte, Gilbert in Journal of the Operational Research Society (2012), 63 The goal of this paper is the development of a new mixed integer linear program designed for optimally loading a set of containers and pallets into a compartmentalised cargo aircraft. It is based on realworld problems submitted by a professional partner. This model takes into account strict technical and safety constraints. In addition to the standard goal of optimally positioning the centre of gravity, we also propose a new approach based on the moment of inertia. This double goal implies an increase in aircraft efficiency and a decrease in fuel consumption. Cargo loading generally remains a manual, or at best a computer assisted, and time consuming task. A fully automatic software was developed to quickly compute optimal solutions. Experimental results show that our approach achieves better solutions than manual planning, within only a few seconds. RelaxMCD: smooth optimisation for the Minimum Covariance Determinant estimator Schyns, Michael; Haesbroeck, Gentiane; Critchley, Frank in Computational Statistics & Data Analysis (2010), 54(4), 843857 The Minimum Covariance Determinant (MCD) estimator is a highly robust procedure for estimating the center and shape of a high dimensional data set. It consists of determining a subsample of h points out of n which minimizes the generalized variance. By definition, the computation of this estimator gives rise to a combinatorial optimization problem, for which several approximative algorithms have been developed. Some of these approximations are quite powerful, but they do not take advantage of any smoothness in the objective function. In this paper, focus is on the approach outlined in a general framework in Critchley et al. (2009) and which transforms any discrete and high dimensional combinatorial problem of this type into a continuous and lowdimensional one. The idea is to build on the general algorithm proposed by Critchley et al. (2009) in order to take into account the particular features of the MCD methodology. More specifically, both the adaptation of the algorithm to the specific MCD target function as well as the comparison of this “specialized” algorithm with the usual competitors for computing MCD are the main goals of this paper. The adaptation focuses on the design of “clever” starting points in order to systematically investigate the search domain. Accordingly, a new and surprisingly efficient procedure based on the well known kmeans algorithm is constructed. The adapted algorithm, called RelaxMCD, is then compared by means of simulations and examples with FASTMCD and the Feasible Subset Algorithm, both benchmark algorithms for computing MCD. As a byproduct, it is shown that RelaxMCD is a general technique encompassing the two others, yielding insight about their overall good performance. A relaxed approach to combinatorial problems in robustness and diagnostics Critchley, Frank; Schyns, Michael; Haesbroeck, Gentiane; Fauconnier, Cécile; Lu, Guobing; Atkinson, Richard A; Wang, Dong Quian in Statistics and Computing (2010), 20(1), 99115 A range of procedures in both robustness and diagnostics require optimisation of a target functional over all subsamples of given size. Whereas such combinatorial problems are extremely difficult to solve exactly, something less than the global optimum can be ‘good enough’ for many practical purposes, as shown by example. Again, a relaxation strategy embeds these discrete, highdimensional problems in continuous, lowdimensional ones. Overall, nonlinear optimisation methods can be exploited to provide a single, reasonably fast algorithm to handle a wide variety of problems of this kind, thereby providing a certain unity. Four running examples illustrate the approach. On the robustness side, algorithmic approximations to minimum covariance determinant (MCD) and least trimmed squares (LTS) estimation. And, on the diagnostic side, detection of multiple multivariate outliers and global diagnostic use of the likelihood displacement function. This last is developed here as a global complement to Cook’s (in J. R. Stat. Soc. 48:133–169, 1986) local analysis. Appropriate convergence of each branch of the algorithm is guaranteed for any target functional whose relaxed form is—in a natural generalisation of concavity, introduced here—‘gravitational’. Again, its descent strategy can downweight to zero contaminating cases in the starting position. A simulation study shows that, although not optimised for the LTS problem, our general algorithm holds its own with algorithms that are so optimised. An adapted algorithm relaxes the gravitational condition itself. A robust heuristic for the optimal selection of a portfolio of stocks Schyns, Michael in International Journal of Operational Research (2010), 9(3), 258271 This paper introduces a new optimization heuristic for the robustification of critical inputs under consideration in many problems. It is shown that it allows to improve significantly the quality and the stability of the results for two classical financial problems, i.e. the Markowitz' portfolio selection problem and the computation of the financial beta. Focus here is on the robust Minimum Covariance Determinant (MCD) estimator which can easily be substituted to the classical estimators of location and scatter. By definition, the computation of this estimator gives rise to a combinatorial optimization problem. We present a new heuristic, called 'RelaxMCD', which is based on a relaxation of the problem to the continuous space. The utility of this approach and the performance of our heuristic, with respect to other competitors, are illustrated through extensive simulations. Optimal selection of a portfolio of options under ValueatRisk constraints: a scenario approach Schyns, Michael; Crama, Yves; Hübner, Georges in Annals of Operations Research (2010), 181 This paper introduces a multiperiod model for the optimal selection of a financial portfolio of options linked to a single index. The objective of the model is to maximize the expected return of the portfolio under constraints limiting its ValueatRisk. We rely on scenarios to represent future security prices. The model contains several interesting features, like the consideration of transaction costs, bidask spreads, arbitragefree option pricing, and the possibility to rebalance the portfolio with options introduced at the start of each period. The resulting mixed integer programming model is applied to realistic test instances involving options on the S&P500 index. In spite of the large size and of the numerical difficulty of this model, nearoptimal solutions can be computed by a standard branchandcut solver or by a specialized heuristic. The structure and the financial features of the selected portfolios are also investigated. Grafting Information in Scenario Trees: Application to Option Prices Schyns, Michael; Crama, Yves; Hübner, Georges Eprint/Working paper (2005) The high level of sophistication in portfolio management modeling techniques often goes along with very large output sensitivity to parameter choices. As a potential solution to this problem, this paper proposes a consistent and flexible methodology to represent the distribution of future values of a portfolio through scenario trees. This methodology relies on the information contained in current option prices in order to generate the probability density function of future returns. This density function can be used, in turn, to generate scenario trees . As an illustration, a tree of scenarios based on S&P500 options is built and then used to compute arbitragefree option prices. The approach preserves information embedded in options prices and is able to provide very accurate values for outofsample options. The high level of numerical accuracy of the framework is reproduced on different samples. The scenario tree approach also provides stable pricing results when confronted with the passage of time. The results derived from our model are comparable to those obtained from Rubinstein’s [1994] methodology, although both models fulfill different objectives. Simulated annealing for complex portfolio selection problems Crama, Yves; Schyns, Michael in European Journal of Operational Research (2003), 150(3), 546571 This paper describes the application of a simulated annealing approach to the solution of a complex portfolio selection model. The model is a mixed integer quadratic programming problem which arises when Markowitz' classical meanvariance model is enriched with additional realistic constraints. Exact optimization algorithms run into difficulties in this framework and this motivates the investigation of heuristic techniques. Computational experiments indicate that the approach is promising for this class of problems. (C) 2003 Elsevier B.V. All rights reserved. Modelling Financial Data and Portfolio Optimization Problems Schyns, Michael Doctoral thesis (2001) This doctoral dissertation in management science, entitled “Modelling Financial Data and Portfolio Optimization Problems”, consists of two independent parts, whose unifying theme is the construction and solution of mathematical programming models motivated by portfolio selection problems. As such, this work is located at the interface of operations research and of finance. It draws heavily on techniques and theoretical results originating in both disciplines. The first part of the dissertation (Chapter 2) deals with an extension of Markowitz model and takes into account some of the sideconstraints faced by a decisionmaker when composing an investment portfolio, viz. lower and upper bounds on the quantities traded, and upper bounds on the number of assets included in the portfolio. We focus on the algorithmic difficulties raised by this model and we describe an original simulated annealing heuristic for its solution. The second (and largest) part of the thesis deals with a new multiperiod model for the optimization of a portfolio of options linked to a single index (Chapters 410). The objective of the model is to maximize the expected return of the portfolio under constraints limiting its valueatrisk. The model contains several interesting features, like the possibility to rebalance the portfolio with options introduced at the start of each period, explicit consideration of transaction costs, realistic pricing of options, consideration of advanced probability models to represent the future, etc. Some deep theoretical results from the financial literature are exploited in order to enrich the model and to extend its applicability. In particular, several available schemes for the generation of scenarios and for option pricing have been critically examined, and the most appropriate ones have been implemented. Furthermore, several optimization approaches (heuristic or exact procedures) have also been developed, implemented and tested. The models investigated in the dissertation bear on very different portfolio problems, draw on separate streams of scientific literature, and are handled by distinct algorithmic techniques. Therefore, the corresponding parts of the dissertation are fully independent, and each part contains its own specific introduction and literature review. Les réseaux de neurones: principes et applicatioàn à la détection financière des faillites Schyns, Michael Eprint/Working paper (1997) Prétraitement de données en reconnaissance de formes par RNA Schyns, Michael Master's dissertation (1993) Les réseaux de neurones n'ont malheureusement pas que des avantages. Leur taille croît avec la quantité et la complexité des données à traiter. Or le principal inconvénient du RNA est lié à sa complexité. Plus elle est grande, plus le RNA sera difficile et coûteux à implémenter physiquement et plus son temps d'apprentissage sera grand. Pour résoudre ce problème, une solution est de prétraiter les données pour diminuer leur taille. Il nous a dès lors été demandé d'analyser trois méthodes de compression applicables au traitement par RNA : la méthode de KarhunenLoève, la méthode LPC (Linear Predictive Coding) et la méthode NLPCA (Non Linear Principal Composant Analysis). Nous en avons ajouté une : la méthode LSP (Line Spectral Pair). Les chapitres qui y sont consacrés tenteront de convaincre le lecteur de l'efficacité de ces procédés. Nous déterminerons également comment choisir la méthode à utiliser pour un problème de classification posé à un RNA. Books and chapters of booksProceedings of the 12th Workshop on Models and Algorithms for Planning and Scheduling Problems MarchettiSpaccamela, Alberto; Crama, Yves; Goossens, Dries; Leus, Roel; Schyns, Michael; Spieksma, Frits Book published by KU Leuven (2015) This volume contains abstracts of talks presented at the 12th Workshop on Models and Algorithms for Planning and Scheduling Problems (MAPSP 2015), held from June 8 to June 12, 2015, in La RocheenArdenne, Belgium. MAPSP is a biennial workshop dedicated to all theoretical and practical aspects of scheduling, planning, and timetabling. The abstracts in this volume include 5 invited talks by Onno Boxma, Michel Goemans, WillemJan van Hoeve, Rolf Niedermeier, and Stephan Westphal, plus 88 contributed talks. Automatic Cargo Load Planning: Special shipments Limbourg, Sabine; Schyns, Michael in Cornelis, Eric (Ed.) Proceedings of the BIVECGIBET Transport Research Day 2011 (2011) The aircraft loading problem is a realworld combinatorial optimisation problem highly constrained. Indeed, loading the aircraft so the gross weight is less than the maximum allowable is not enough. This weight must be distributed to keep the centre of gravity (CG) within specified limits. Moreover, an aircraft has usually several cargo compartments with specific contours and structural limitations such as floor loading, combined load limits and cumulative load limitations. Finally, some shipments are particularly restrictive to transport, like dangerous goods, live animals and perishable goods. This paper is concerned with the incorporation of these latter constraints in a mixed integer linear program for the problem of loading a set of Unit Loading Devices (ULDs) and bulk into an aircraft. Experimental results for real data sets show that the model achieves better balanced solutions in only a few seconds compared to the solution obtained by load masters. Alternative to the MeanVariance Asset Allocation Analysis: A Scenario Methodology for Portfolio Selection Schyns, Michael; Hübner, Georges; Crama, Yves in Gregoriou, Greg N. (Ed.) Stock Market Volatility (2009) This paper introduces a new methodology to optimize the allocation of financial assets. The objective of the model is to maximize the expected return of the portfolio under constraints limiting its ValueatRisk. The assets could consist in stocks as well as options. We rely on a flexible scenario tree approach to represent the future prices. In order to reduce the number of leaves and maintain the model tractable, stocks prices are obtained through the Fama & French empirical asset pricing model. Experiments on historical data are performed to illustrate the method and show the performance of the approach. Different strategies are compared: considering various market distributions, several factor models and a few portfolio hypothesis. Visualizing Statistical Models and Concepts Farebrother, Robert W.; Schyns, Michael Book published by Marcel Dekker, Inc (2002) This text/reference examines classic algorithms, geometric diagrams, and mechanical principles for enhanced visualization of statistical estimation procedurs and mathematical concepts in physics, engineering, and computer programming  stressing the role of geometric and mechanical representations in the design and generation of numericl models for applications in physical science. Visualizing Statistical Models and Concepts provides methods to determine the position of a multivariate location parameter ... offers techniques to fit a plane or curved surface to multivariate data ... considers geomechanical dynamics for linear and non linear programming ... discusses mathematical approaches to estimate and predict potential energy, force, and strain of system components...and analyzes data sets generated by transitive and nontransitive pairwise preference orderings. Théorie stochastique de la décision d'investissement Justens, Daniel; Schyns, Michael Book published by De Boeck & Larcier (1997) La complexité, l'imprévisibilité croissantes de l'environnement et de la structure interne des entreprises contraignent le gestionnaire à tendre vers une description probabiliste de l'univers. Le passage de la théorie classique, paramétrique, incluant la variabilité des données dans un contexte déterministe et discret, à la théorie stochastique continue, se fait progressivement. Chaque étape est illustrée au moyen d'exemples concrets. Un logiciel (sur CDROM) permet la résolution immédiate des problèmes propres à l'utilisateur. ConferencesEarly detection of university students with potential difficulties Hoffait, AnneSophie; Schyns, Michael Conference (2017, July) Using data mining methods, this paper presents a new means of identifying freshmen's profiles likely to face major difficulties to complete their first academic year. We aim at early detection of potential failure using student data available at registration, i.e. school records and environmental factors, with a view to timely and efficient remediation and/or study reorientation. We adapt three data mining methods, namely random forest, logistic regression and artificial neural network algorithms. We design algorithms to increase the accuracy of the prediction when some classes are of major interest. These algorithms are context independent and can be used in different fields. They rely on a dynamic split of the observations into subclasses during the training process, so as to maximize an accuracy criterion. Four classes are so built: high risk of failure, risk of failure, expected success or high probability of success. Real data pertaining to undergraduates at the University of Liège (Belgium), illustrates our methodology. With our approach, we are now able to identify with a high rate of confidence (90%) a subset of 12.2% of students facing a very high risk of failure, almost the quadruple of those identified with a nondynamic approach. By testing some confidence levels, our approach makes it possible to rank the students by levels of risk and a sensitivity analysis allows us to find out why some students are likely to encounter difficulties. A best fit heuristic for the threedimensional Bin Packing Problem with practical constraints from an air transportation application Paquay, Célia; Limbourg, Sabine; Schyns, Michael in Proceedings of the BIVECGIBET Transport Research Days 2017 (2017) This work considers the threedimensional Multiple Bin Size Bin Packing Problem which consists in packing a set of cuboid boxes into containers of various shapes, while minimising unused space. The aim of the present work is to find good initial solutions in short computational times. In this purpose, a fast best fit heuristic able to manage the different constraints to be met in transportation is developed. The heuristic is split into two distinct phases. The first phase deals with the packing of boxes into identical bins using an extension of the Extreme Points rule. During this phase, the fragility, stability and orientation of the boxes are taken into account, as well as the special shape of the bins and their weight capacity. The second phase considers the multiple types of available bins. If necessary, the best loading pattern identified is enhanced with respect to weight distribution in post processing. After parametrisation, computational experiments are performed on data sets specially designed for this application. The heuristic requires very short computational times to achieve promising results. Early detection of university students in potential difficulty Hoffait, AnneSophie; Schyns, Michael Conference (2016, July) Accounting for price endogeneity in airline itinerary choice models Lurkin, Virginie; Garrow, Laurie; Higgins, Matthew; Newman, Jeffrey; Schyns, Michael Conference (2016, May 20) This study formulates an itinerary choice model that is consistent with those used by industry and corrects for price endogeneity using a control function that uses several types of instrumental variables. We estimate our models using database of more than 3 million tickets provided by the Airlines Reporting Corporation. Results based on Continental U.S. markets for May 2013 departures show that models that fail to account for price endogeneity overestimate customers’ value of time and result in biased price estimates and incorrect pricing recommendations. Extensions to advanced discrete choice models show the importance of accounting for interalternative substitution for products that share similar departure times. Estimation of Airline Itinerary Choice Models Using Disaggregate Ticket Data Lurkin, Virginie; Garrow, Laurie A.; Higgins, Matthew J.; Schyns, Michael in 55th AGIFORS Annual Proceedings 2015 (2015, August 29) Airline itinerary choice models support many multimillion dollar decisions, i.e., they are used to evaluate potential route schedules. Classic models suffer from major limitations, most notably they use average fare information but to not correct for price endogeneity. We use a novel database of airline tickets to estimate itinerary choice models using detailed fare data and compare these to classic itinerary choice models that use aggregate fare information but correct for price endogeneity. A best fit decreasing algorithm for the three dimensional bin packing problem Paquay, Célia; Schyns, Michael; Limbourg, Sabine Conference (2015, July) What the heck is Revenue Management Lurkin, Virginie; Schyns, Michael; Garrow, Laurie Scientific conference (2015, May 08) The airline industry changed dramatically in 1978 when it became deregulated. Operations research analysts played a critical role after deregulation by developing algorithms and decisionsupport systems designed to help airlines to maximize their revenue. More than thirtyfive years after deregulation, the airline industry is faced with new challenges. The increased use of the Internet as the major distribution channel and the increased market penetration of low cost carriers have led to an increasing interest in using discrete choice models to model air travel demand as the collection of individual's decisions. What is the impact of ticketlevel fare information on classic itinerary choice models ? Lurkin, Virginie; Garrow, Laurie; Schyns, Michael Scientific conference (2015, March 25) I have been invited by Prof. Dr. Catherine Cleophas as an external guest to the "Revenue Management Colloquium" in AixlaChapelle from March 25 to March 26. The aim is to present my own doctoral project and discuss the presentations of other PhD students Several constructive heuristics for the three dimensional Multiple Bin Size Bin Packing Problem with air transportation constraints Paquay, Célia; Limbourg, Sabine; Oliveira, José Fernando; Schyns, Michael Conference (2015, March) Early detection of university students in potential difficulty Hoffait, AnneSophie; Schyns, Michael Conference (2015, February) Rate of success in the first year at University in Belgium is very low regarding other foreign universities. The University of Liege, as other Universities, has already taken different initiatives. But by early identifying students who have a high probability to face difficulties if nothing is done, the Universities might develop adapted methods to attack the problem with more emphasis where it is more needed and when it is still possible. Thus we want to develop a decision tool able to identify these students to help them. For that, we consider three standard datamining methods: logistic regression, artificial neural networks and decision trees and focus on early detection, i.e. before starting at the University. Then, we suggest to adapt these three methods as well as the classification framework in order to increase the probability of correct identification of the students. In our approach, we do not restrict the classification to two extreme classes, e.g. failure or success, but we create subcategories for different levels of confidence: high risk of failure, risk of failure, expected success or high probability of success. The algorithms are modified accordingly and to give more weight to the class that really matters. Note that this approach remains valid for any other classification problems for which the focus is on some extreme classes; e.g. fraud detection, credit default... Finally, simulations are conducted to measure the performances of the three methods, with and without the suggested adaptation. We check if the factors of success/failure we can identify are similar to those reported in the literature. We also make a ``whatif sensitivity analysis''. The goal is to measure in more depth the impact of some factors and the impact of some solutions, e.g., a complementary training or a reorientation. The Airline Container Loading Problem with Pickup and Delivery Lurkin, Virginie; Schyns, Michael Conference (2014, November 09) We address the problem of allocating containers into predefined positions of a carrier, in this case aircraft, under several realistic structural and safety constraints. The originality of our approach is to allow multitrips with pickup and delivery at some intermediate locations. The objective is to minimize the economic and environmental costs including the impact of the intermediate operations. We resort to an integer linear model. Numerical experiments have been performed using a standard B\&C library. Heuristics are developed to speed up the process. Early detection of university students in potential difficulty : a case study Hoffait, AnneSophie; Schyns, Michael Conference (2014, November) Rate of success in the first year at University in Belgium is very low regarding other foreign universities. The University of Liege, as other Universities, has already taken different initiatives. But by early identifying students who have a high probability to face difficulties if nothing is done, the Universities might develop adapted methods to attack the problem with more emphasis where it is more needed and when it is still possible. Thus we want to develop a decision tool able to identify these students to help them. For that, we consider three standard datamining methods: logistic regression, artificial neural networks and decision trees and focus on early detection, i.e. before starting at the University. Then, we suggest to adapt these three methods as well as the classification framework in order to increase the probability of correct identification of the students. In our approach, we do not restrict the classification to two extreme classes, e.g. failure or success, but we create subcategories for different levels of confidence: high risk of failure, risk of failure, expected success or high probability of success. The algorithms are modified accordingly and to give more weight to the class that really matters. Note that this approach remains valid for any other classification problems for which the focus is on some extreme classes; e.g. fraud detection, credit default... Finally, simulations are conducted to measure the performances of the three methods, with and without the suggested adaptation. We check if the factors of success/failure we can identify are similar to those reported in the literature. We also make a ``whatif sensitivity analysis''. The goal is to measure in more depth the impact of some factors and the impact of some solutions, e.g., a complementary training or a reorientation. The Airline Container Loading Problem with Pickup and Delivery Lurkin, Virginie; Schyns, Michael in Airline Group of the Intl Federation of Operational Research Soc ( AGIFORS ) (Ed.) 54th AGIFORS Annual Proceedings 2014 (2014, October 19) The present paper looks into the problem of optimizing the loading of a set of containers and pallets into cargo aircraft serving multiple airports. Due to the pickup and delivery operations occurring at intermediate airports, this problem is simultaneously a weight and balance problem and a sequencing problem. Our objective is to minimize fuel and handling operations costs. This problem is shown to be NPhard. We resort to a mixed integer linear program. On the basis of a professional partner's realworld data, TNT Airways, we perform numerical experiments using a standard B&C library. This approach yields better solutions than traditional manual planning, which results in substantial cost savings. A 'price balance statistic' for optimizing pricing strategies: a better estimation of elasticities and crosselasticities Lurkin, Virginie; Schyns, Michael; Garrow, Laurie A.; Jacobs, Timothy L. Conference (2014, May 16) Demand forecasting, price optimization and capacity controls form three major tools of revenue management. Over the past few decades, each discipline has generated a great deal of research but has typically been studied separately from the others. Yet, better understanding their relationship gives an airline the opportunity to increase its profitability. In prior work, Tim Jacobs and colleagues introduced a macrolevel metric known as the ‘Price Balance Statistic (PBS)’ for evaluating the quality of a given pricing strategy and guiding a search algorithm to identify an optimal alignment between pricing structure, scheduled capacity and RM controls using marginal revenue principles. The aim of our work is to incorporate additional modeling improvements to the PBS. The current model formulation uses price elasticity as input parameters and assumes perfect independency between the different fare classes. However, in reality, a passenger demand fluctuates between classes based on differences in prices. We propose to use instrumented variable linear regression methods to obtain parameter estimates for price elasticities and crosselasticities. This modification incorporates accurate price elasticities but also the impact of a change in one fare class on another through the crosselasticities. A 'price balance statistic' for optimizing pricing strategies: a better estimation of elasticities and crosselasticities Lurkin, Virginie; Schyns, Michael; Garrow, Laurie A.; Jacobs, Timothy L. Conference (2014, May 15) Demand forecasting, price optimization and capacity controls form three major tools of revenue management. Over the past few decades, each discipline has generated a great deal of research but has typically been studied separately from the others. Yet, better understanding their relationship gives an airline the opportunity to increase its profitability. In prior work, Tim Jacobs and colleagues introduced a macrolevel metric known as the ‘Price Balance Statistic (PBS)’ for evaluating the quality of a given pricing strategy and guiding a search algorithm to identify an optimal alignment between pricing structure, scheduled capacity and RM controls using marginal revenue principles. The aim of our work is to incorporate additional modeling improvements to the PBS. The current model formulation uses price elasticity as input parameters and assumes perfect independency between the different fare classes. However, in reality, a passenger demand fluctuates between classes based on differences in prices. We propose to use instrumented variable linear regression methods to obtain parameter estimates for price elasticities and crosselasticities. This modification incorporates accurate price elasticities but also the impact of a change in one fare class on another through the crosselasticities. A relaxand fix heuristic for the threedimensional binpacking model with transportation constraints Paquay, Célia; Schyns, Michael; Limbourg, Sabine Conference (2014, January 30) Early detection of university students in potential difficulty Hoffait, AnneSophie; Schyns, Michael Conference (2014, January) Rate of success in the first year at University in Belgium is very low regarding other foreign universities. The University of Liege, as other Universities, has already taken different initiatives. But by early identifying students who have a high probability to face difficulties if nothing is done, the Universities might develop adapted methods to attack the problem with more emphasis where it is more needed and when it is still possible. Thus we want to develop a decision tool able to identify these students to help them. For that, we consider three standard datamining methods: logistic regression, artificial neural networks and decision trees and focus on early detection, i.e. before starting at the University. Then, we suggest to adapt these three methods as well as the classification framework in order to increase the probability of correct identification of the students. In our approach, we do not restrict the classification to two extreme classes, e.g. failure or success, but we create subcategories for different levels of confidence: high risk of failure, risk of failure, expected success or high probability of success. The algorithms are modified accordingly and to give more weight to the class that really matters. Note that this approach remains valid for any other classification problems for which the focus is on some extreme classes; e.g. fraud detection, credit default... We check if the factors of success/failure we can identify are similar to those reported in the literature. We also make a ``whatif sensitivity analysis''. The goal is to measure in more depth the impact of some factors and the impact of some solutions, e.g., a complementary training or a reorientation. A constructive heuristic for the three dimensional Bin Packing Problem with transportation constraints Paquay, Célia; Schyns, Michael; Limbourg, Sabine Conference (2014) The aim of this work is to propose a RelaxAndFix heuristic to build a good initial solution to the 3D BPP. First, a mathematical formulation has been developed taking into account several types of constraints such as the stability and fragility of the boxes to pack, their possibility to rotate, the weight distribution inside the bins and their special shapes. Since this model contains a lot of integer variables, we have decided to apply the RelaxandFix method. We have selected several sets of variables to be the branching variables and carried out some tests. A relaxand fix heuristic for the threedimensional binpacking model with transportation constraints Paquay, Célia; Schyns, Michael; Limbourg, Sabine in 11th ESICUP Meeting proceedings (2014) The Airline Container Loading Problem with Pickup & Delivery and Multi Doors Schyns, Michael; Lurkin, Virginie in 53rd AGIFORS Annual Proceedings 2013: Annual Symposium and Study Group Meeting (2013, August) We address the problem of allocating containers into predefined positions of a carrier, in this case aircraft, under several realistic structural and safety constraints, including the management of several doors. The originality of our approach is to allow multitrips with pickup and delivery at some intermediate locations. The objective is to minimize the economic and environmental costs including the impact of the intermediate operations. We resort to an integer linear model. Numerical experiments have been performed using a standard B&C library. The cost impact is measured. The Airline Group of the International Federation of Operational Research Societies, composed of professionals and academics, has awarded us the price of the "Best technical presentation" for this work. The Airline Container Loading Problem with Pickup and Delivery Lurkin, Virginie; Schyns, Michael Conference (2013, July 02) We address the problem of allocating containers into predefined positions of a carrier, in this case aircraft, under several realistic structural and safety constraints. The originality of our approach is to allow multitrips with pickup and delivery at some intermediate locations. The objective is to minimize the economic and environmental costs including the impact of the intermediate operations. We resort to an integer linear model. Numerical experiments have been performed using a standard B\&C library. Heuristics are developed to speed up the process. A branch and price approach for an airport vehicle routing problem Schyns, Michael Conference (2013, July) This project has been initiated by a main European freight airport. The goal is to optimize the aircraft refueling process which relies on a given set of trucks. The underlying process can be defined as a vehicle routing problem with capacity and time windows. We resort to a branch and price approach for which we first analyze the impact of different parameters on the performance of the algorithm. Due to the stochastic nature of the demand and time windows in this context, we are also working on a priori (split delivery) and a posteriori (efficient reoptimization) measures. An exact formulation for the threedimensional binpacking with transportation constraints Paquay, Célia; Schyns, Michael; Limbourg, Sabine Conference (2013, July) Automatic Aircraft Cargo Load Planning with Pickup and Delivery Lurkin, Virginie; Schyns, Michael Conference (2013, March 15) This research aims to develop a new mixed integer linear program to solve the containers assignment problem when pickup and deliveries are considered. Given a pool of ULDs and an aircraft with multiple destinations, we want to obtain a loading plan determining at which positions the ULDs must be assigned in order to minimize simultaneous the quantity of fuel consumed and the number of rehandles. The loading plan should also ensure a number of structural, safety and manoeuvrability constraints. The model has been tested on real instances and provides encouraging results. Automatic Aircraft Cargo Load Planning with Pickup and Delivery Lurkin, Virginie; Schyns, Michael Conference (2013, February 07) This research aims to develop a new mixed integer linear program to solve the containers assignment problem when pickup and deliveries are considered. Given a pool of ULDs and an aircraft with multiple destinations, we want to obtain a loading plan determining at which positions the ULDs must be assigned in order to minimize simultaneous the quantity of fuel consumed and the number of rehandles. The loading plan should also ensure a number of structural, safety and manoeuvrability constraints. The model has been tested on real instances and provides encouraging results. Automatic Cargo Load Planning: Special shipments Kleyntssens, Thomas; Limbourg, Sabine; Schyns, Michael in ILS 2012 Proceedings (2012, August 28) The aircraft loading problem is a realworld combinatorial optimisation problem highly constrained. Indeed, loading the aircraft so the gross weight is less than the maximum allowable is not enough. This weight must be distributed to keep the centre of gravity within specified limits. Moreover, an aircraft has usually several cargo compartments with specific contours and structural limitations such as floor loading, combined load limits and cumulative load limitations. Finally, some shipments are particularly restrictive to transport, like dangerous goods, live animals and perishable goods. This paper is concerned with the incorporation of these latter constraints in a mixed integer linear program for the problem of loading a set of Unit Loading Devices and bulk into an aircraft. Experimental results show that our method achieves optimal solutions within only few seconds. Three dimensional Bin Packing Problem applied to air cargo Paquay, Célia; Schyns, Michael; Limbourg, Sabine in ILS 2012 Proceedings (2012, August 26) Deciding whether a set of three dimensional boxes can be packed into a container is a NPhard problem. Mathematical models have been developed, however, only few studies take into account constraints encountered in realworld applications such as the stability or the fragility of the cargo. Moreover, despite the importance of this issue in air transport, the literature is almost silent on constraints related to the distribution of the weight inside a container. This paper is concerned with the formulation of the three dimensional palletization which includes the main constraints met in the air cargo industry. Optimisation 3D du chargement de conteneurs pour le transport aérien Paquay, Célia; Limbourg, Sabine; Schyns, Michael in LigéRO (Ed.) Proceedings ROADEF 2013 (2012, April 11) De nos jours, décider comment remplir des conteneurs avec des colis est une activité courante aussi bien dans le domaine du transport routier qu’aérien. Ce type de question est apparentée aux problèmes de BinPacking(BPP) en recherche opérationnelle. En termes économiques, ces opérations doivent mener à une solution qui satisfait de nombreuses contraintes physiques, être réalisées rapidement et de sorte à maximiser certains critères. Typiquement, on tentera de charger un maximum de colis dans un minimum de conteneurs pour réduire les coûts. Chargement de marchandises dans un avion cargo: le cas des marchandises nécessitant des précautions particulières Schyns, Michael; Limbourg, Sabine; Kleyntssens, Thomas Conference (2012, April) D'une part, les entreprises de transport aérien ont acheminé en 2010 plus d’un tiers de la valeur des exportations mondiales. D'autre part, le chargement des avions est une opération complexe soumise à de nombreuses contraintes et peu d'outils sont disponibles pour aider les loadmasters à trouver la meilleure disposition des conteneurs dans les avions. Limbourg, Schyns et Laporte (2011) ont proposé un modèle à variables entières pour traiter les problèmes élémentaires. Notre travail est une extension de ces travaux. Nous considérons des chargements spéciaux qui impliquent des précautions particulières (produits dangereux, animaux, produits réfrigérés, aliments périssables, ...) ainsi que le transport de marchandises de plus grande taille. L’ajout de ces deux types de contraintes se justifie par la grande fréquence de ces situations dans des problèmes réels rencontrés par nos partenaires industriels. Le problème résultant est très complexe et nous proposons un outil pour le résoudre. An Integer Programming model for air transport of hazardous and special shipments Kleyntssens, Thomas; Limbourg, Sabine; Schyns, Michael Conference (2012, February 02) The aircraft loading problem is a realworld combinatorial optimisation problem highly constrained. This weight must be distributed to keep the centre of gravity within speci ed limits. Moreover, an aircraft has usually several cargo compart ments with speci c contours and structural limitations such as oor loading, com bined load limits and cumulative load limitations. Finally, some shipments are par ticularly restrictive to transport, like dangerous goods, live animals and perishable goods. This paper is concerned with the incorporation of these latter constraints in a mixed integer linear program for the problem of loading a set of Unit Load ing Devices and bulk into an aircraft. Experimental results show that our method achieves optimal solutions within only few seconds. Three dimensional Bin Packing Problem applied to air transport Paquay, Célia; Schyns, Michael; Limbourg, Sabine Conference (2012, February 02) Packing boxes into containers is a daily process in many di erent elds and especially in transport. However, the particular case of air transport brings some new constraints such as the stability or the fragility of the cargo. The distribution of the weight has also to be considered. Moreover, this special case also brings some data such as the dimensions of the possible containers, called Unit Load Devices. This paper is concerned with the formulation of the three dimensional palletization which includes the main constraints met in the air cargo industry. It proposes a integer linear program for this combinatorial optimization problem. Three Dimensional Bin Packing Problem applied to air cargo Paquay, Célia; Schyns, Michael; Limbourg, Sabine Scientific conference (2011, December 15) Deciding whether a set of three dimensional boxes can be packed into a container is a NPhard problem. Mathematical models have been developed, however, only few studies take into account constraints encountered in realworld applications such as the stability or the fragility of the cargo. Moreover, despite the importance of this issue in air transport, the literature is almost silent on constraints related to the distribution of the weight inside a container. This paper is concerned with the formulation of the three dimensional palletization which includes the main constraints met in the air cargo industry. Chargement d’un avion cargo : le cas des marchandises nécessitant des précautions particulières Kleyntssens, Thomas; Limbourg, Sabine; Schyns, Michael Conference (2011, December 15) Le problème du chargement de marchandises dans un avion cargo est soumis à des contraintes strictes de sécurité. C’est un problème d’optimisation combinatoire d’une importance cruciale pour les compagnies aériennes. En effet, un mauvais chargement diminue l’efficacité d’un avion et impose des tensions importantes sur sa structure qui peuvent entraîner la destruction d’équipements de haute valeur, voire la perte de vies. De plus, certains colis spéciaux ont des contraintes très restrictives. C’est le cas notamment des produits dangereux, des animaux, des marchandises alimentaires et périssables. En plus des incompatibilités entre certains produits, il faut tenir compte des interactions possibles sur des équipements de l’avion. Par exemple, des produits émettant des émissions magnétiques doivent être placés de manière à ne pas avoir des effets néfastes avec les instruments de navigation. Dans cet article, nous proposons d’incorporer ces contraintes liées aux colis nécessitant des précautions particulières dans un programme d’optimisation linéaire mixte du problème de chargement d’un ensemble de Unit Loading Devices et de vrac dans un avion. Les résultats obtenus à partir de données réelles montrent que notre méthode permet d’obtenir des solutions optimales en seulement quelques secondes. Automatic Cargo Load Planning: Special shipments Limbourg, Sabine; Schyns, Michael Scientific conference (2011, May 25) The aircraft loading problem is a realworld combinatorial optimisation problem highly constrained. Indeed, loading the aircraft so the gross weight is less than the maximum allowable is not enough. This weight must be distributed to keep the centre of gravity (CG) within specified limits. Moreover, an aircraft has usually several cargo compartments with specific contours and structural limitations such as floor loading, combined load limits and cumulative load limitations. Finally, some shipments are particularly restrictive to transport, like dangerous goods, live animals and perishable goods. This paper is concerned with the incorporation of these latter constraints in a mixed integer linear program for the problem of loading a set of Unit Loading Devices (ULDs) and bulk into an aircraft. Experimental results for real data sets show that the model achieves better balanced solutions in only a few seconds compared to the solution obtained by load masters. Planification automatique de chargements d'avions cargo Limbourg, Sabine; Schyns, Michael Scientific conference (2011, April 18) Identification du profil du consommateur par datamining Schyns, Michael Scientific conference (2011, April) Automatic Cargo Load Planning Limbourg, Sabine; Schyns, Michael; Laporte, Gilbert in Proceedings (2011, March 02) The goal of this paper is the development of a new mixed integer linear pro gram designed for optimally loading a set of containers and pallets into a compartmentalised cargo aircraft. It is based on realworld problems submitted by a professional partner. This model takes into account strict technical and safety constraints. In addition to the standard goal of optimally positioning the centre of gravity, we also propose a new approach based on the moment of inertia. This double goal implies an increase in aircraft efficiency and a decrease in fuel consumption. Cargo loading generally remains a manual, or at best a computer assisted, and time consuming task. A fully automatic software was developed to quickly compute optimal solutions. Experimental results show that our approach achieves better solutions than manual planning, within only a few seconds. Robust Portfolio Selection Schyns, Michael in JSM Proceedings, Statistical Computing Section (2008, November) In many financial problems, small variations in some inputs may result in big changes in the outputs. In this talk, we consider the problem of portfolio selection as suggested by Markowitz. This model relies on a covariance matrix usually estimated using historical returns of the assets under consideration. Gross error in these returns or atypical events occurring in the past could lead to different portfolios with quite different expected returns. Defining methods that do not depend too much on these atypical data is the aim of robust statistics. We will show that some techniques developed in that field are worth applying in our context. More precisely, the covariance matrix of historical data will be estimated with the Minimum Covariance Determinant estimator, computed with a 'smooth' algorithm. This robust Markowitz methodology will be illustrated on real financial data. Solving the mTSP Problem with Stochastic or Time Dependent Demands Louveaux, François; Schyns, Michael in Opasanon, S.; MillerHooks, E. (Eds.) Proceedings of TRISTAN V (Triennial Symposium on Transportation Analysis) (2004, June) There are many examples of problems in transportation where some elements are uncertain. In the distribution of goods as well as systems responding to calls for emergency, demands typically occur in a random fashion. Transportation systems have thus to be created in face of uncertainty about future levels of demands, making strategic decisions difficult to take. Similarly, traffic conditions vary randomly over time and travel routes are usually designed in face of uncertainty about traffic conditions, hence about effective travel times. Stochastic models, i.e. models that take uncertainty explicitly into account, have thus a central role to play in transportation. The case sensitivity function approach to diagnostic and robust computation: a relaxation strategy Critchley, Frank; Schyns, Michael; Haesbroeck, Gentiane; Kinns, D.; Atkinson, R. A.; Lu, G. in Antoch, Jaromir (Ed.) COMPSTAT 2004: Proceedings in Computational Statistics (2004) MiscellaneousDoc'Data: mesure de la persévérance au doctorat à l'Université de Liège Aerts, Stéphanie; Haesbroeck, Gentiane; Schyns, Michael Report (2015) Le Conseil du Doctorat a parmi ses missions celle d’accompagner les doctorants dans leur parcours et de mettre en place des actions en vue de favoriser la réussite. Cependant, pour réellement mesurer l’impact des actions mises en oeuvre dans cette optique, il faut d’abord pouvoir quantiﬁer le taux de réussite. Or, mesurer le taux de réussite dans les études de 3ème cycle n’est pas une démarche évidente. De nombreux raccourcis de raisonnement sont tentants mais mènent à des résultats très variables et incorrects. Le but de ce document est de présenter les diﬃcultés inhérentes au calcul d’un taux de réussite et de proposer quelques procédures appropriées. Dans un premier temps, les chances (ou, d’un point de vue plus formel, les probabilités) de réussite et d'abandon des doctorants sont modélisées en fonction du nombre d’années d’inscription et en tenant compte de caractéristiques du doctorant comme son genre, sa nationalité, son âge au moment de l’inscription, son statut professionnel... L'analyse descriptive et exploratoire est ensuite conﬁrmée par l’application d’une analyse de survie avec risques compétitifs. RAPPORT DE RECHERCHE SUR UNE APPLICATION DE GESTION DE LA COLLABORATION DU SERVICE EXPÉDITION DU CHARGEUR AVEC LE TRANSPORTEUR Pironet, Thierry; Crama, Yves; Arda, Yasemin; Schyns, Michael; Rosso, Vanessa; Kronus, David; Wauthelet, Yves; Lange, JeanCharles; Tran, Vi Report (2009) Dans ce rapport, les échanges d'information d'un donneur d'ordre, d'un transporteur et d'un fournisseur au sein d'une chaîne de distribution sont analysés. Les possibilités de tracking et de tracing de TransLogisTIC sont utilisés pour générer des KPI de performance et un modèle d'optimisation de quais de chargement est décrit dans une version Offline et Online. Informatique en Sciences de Gestion et en Sciences Economiques  Niveau 2 Schyns, Michael Learning material (2009) RAPPORT INTERNE : REVUE DE LITTERATURE SUR LA GESTION DES RESSOURCES RÉUTILISABLES ET DES MÉTHODES D’OPTIMISATION Pironet, Thierry; Crama, Yves; Arda, Yasemin; Schyns, Michael; Rosso, Vanessa; Kronus, David Report (2009) Dans cette revue de la littérature scientifique, on peut trouver une synthèse des thématiques liées aux modèles et aux techniques d'optimisation utilés pour la gestion de ressources réutilisables dans un réseau tels que les containers entre des ports de mers ou des wagons dans un réseau ferroviaire. Les textes fondateurs sont mentionnés dans un ordre historique et un commentaire est fait soit sur le modèle investiqué et ses particularités ou la technique de résolution. RAPPORT DE RECHERCHE SUR L’OPTIMISATION DU ROUTAGE ET DU CHARGEMENT DE VEHICULES. Pironet, Thierry; Crama, Yves; Arda, Yasemin; Schyns, Michael; Rosso, Vanessa; Kronus, David Report (2009) Dans ce rapport confidentiel, un algorithme d'optimisation du chargement de véhicules a été mis au point dans le cadre d'une application avec un partenaire industriel. Academic genealogy: 

Michaël Schyns  Management Information Systems Useful links: University of Liège  HECManagement School  QuantOM  Chaudfontaine Email: M.Schyns 
