Papers by Laetitia Jourdan
2018 IEEE 30th International Conference on Tools with Artificial Intelligence (ICTAI)
Multi-objective optimisation algorithms expose various parameters that have to be tuned in order ... more Multi-objective optimisation algorithms expose various parameters that have to be tuned in order to be efficient. Moreover, in multi-objective optimisation, the correlation between objective functions is known to affect search space structure and algorithm performance. Considering the recent success of automatic algorithm configuration (AAC) techniques for the design of multi-objective optimisation algorithms, this raises two interesting questions: what is the impact of correlation between optimisation objectives on (1) the efficacy of different AAC approaches and (2) on the optimised algorithm designs obtained from these automated approaches? In this work, we study these questions for multi-objective local search algorithms (MOLS) for three well-known bi-objective permutation problems, using two single-objective AAC approaches and one multiobjective approach. Our empirical results clearly show that overall, multi-objective AAC is the most effective approach for the automatic configuration of the highly parametric MOLS framework, and that there is no systematic impact of the degree of correlation on the relative performance of the three AAC approaches. We also find that the best-performing configurations differ, depending on the correlation between objectives and the size of the problem instances to be solved, providing further evidence for the usefulness of automatic configuration of multiobjective optimisation algorithms.
Proceedings of the Genetic and Evolutionary Computation Conference Companion
Mandarine Academy is a major MOOC (Massive Open Online Course) operator with more than 100 active... more Mandarine Academy is a major MOOC (Massive Open Online Course) operator with more than 100 active e-learning platforms in multiple languages and 550K overall users. The company is providing a wide range of content types to help users in their learning process. With thousands of pedagogical online resources in an everyday growing catalog, users can have a hard time finding relevant content. Mandarine Academy is looking to improve the overall user experience while minimizing both information overload and drop rate levels. This paper details the conception and implementation of a multi-objective e-learning recommender system. The proposed approach takes advantage of multiple known implementations like content-based and collaborative filtering among other techniques to generate an initial population of solutions. A custom genetic operator is proposed and compared to classical operators using multiple algorithms (NSGA-II, NSGA-III, SPEA-2, IBEA, and MOEA/D). Using the best configurations found by the tuning process, we showcase the initial results obtained by each approach when applied to real-world data from a public MOOC provided by the company. CCS CONCEPTS • Applied computing → Multi-criterion optimization and decision-making; • Information systems → Collaborative filtering.
2020 IEEE Congress on Evolutionary Computation (CEC)
Classification problems can be modeled as multiobjective optimization problems. MOCA-I is a multi... more Classification problems can be modeled as multiobjective optimization problems. MOCA-I is a multi-objective local search designed to solve these problems, particularly when the data are imbalanced. However, this algorithm has been tuned by hand in order to be efficient on particular datasets. In this paper, we propose a methodology to automatically configure a multi-objective algorithm for solving a supervised partial classification problem. This methodology is based on a multi-objective approach of automatic algorithm configuration and requires a clear definition of the experimental protocol. Therefore, we present a k-fold cross-validation protocol to train and test the configuration model. To the best of our knowledge, it is the first time that multi-objective automatic algorithm configuration is performed on optimization algorithms to solve classification problems. Experimental results on real imbalanced datasets show that our approach can find efficient configurations of MOCA-I with less effort in comparison with the ones found exhaustively by hand.
2020 IEEE Congress on Evolutionary Computation (CEC)
This study proposes a novel multi-objective integer programming model for a collision-free discre... more This study proposes a novel multi-objective integer programming model for a collision-free discrete drone path planning problem. Considering the possibility of bypassing obstacles or flying above them, this study aims to minimize the path length, energy consumption, and the accumulated maximum path risk simultaneously. The static environment is represented as 3D grid cells. Due to the NP-hardness nature of the problem, several state-of-the-art evolutionary multi-objective optimization (EMO) algorithms with customized crossover and mutation operators are applied to find a set of non-dominated solutions. The results show the effectiveness of applied algorithms in solving several generated test cases.
Proceedings of the Genetic and Evolutionary Computation Conference, 2017
Multi-objective local search (MOLS) algorithms are e cient metaheuristics, which improve a set of... more Multi-objective local search (MOLS) algorithms are e cient metaheuristics, which improve a set of solutions by using their neighbourhood to iteratively nd be er and be er solutions. MOLS algorithms are versatile algorithms with many available strategies, rst to select the solutions to explore, then to explore them, and nally to update the archive using some of the visited neighbours. In this paper, we propose a new generalisation of MOLS algorithms incorporating new recent ideas and algorithms. To be able to instantiate the many MOLS algorithms of the literature, our generalisation exposes numerous numerical and categorical parameters, raising the possibility of being automatically designed by an automatic algorithm con guration (AAC) mechanism. We investigate the worth of such an automatic design of MOLS algorithms using MO-ParamILS, a multi-objective AAC con gurator, on the permutation owshop scheduling problem, and demonstrate its worth against a traditional manual design. CCS CONCEPTS • eory of computation → Design and analysis of algorithms; Randomized local search; •Applied computing → Multi-criterion optimization and decision-making;
HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci-entific ... more HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et a ̀ la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. in ria
This paper presents and experiments approaches to solve a new bi-objective routing problem called... more This paper presents and experiments approaches to solve a new bi-objective routing problem called the ring star problem. It consists of locating a simple cycle through a subset of nodes of a graph while optimizing two kinds of cost. The first objective is the minimization of a ring cost that is related to the length of the cycle. The second one is the minimization of an assignment cost from non-visited nodes to visited ones. In spite of its obvious bi-objective formulation, this problem has always been investigated in a single-objective way. To tackle the bi-objective ring star problem, we first investigate different standalone search methods. Then, we propose two cooperative strategies that combines two multiple objective metaheuristics: an elitist evolutionary algorithm and a population-based local search. We apply this new hybrid approaches to well-known benchmark test instances and demonstrate their effectiveness in comparison to non-hybrid algorithms and to state-of-the-art methods.
2016 IEEE Symposium Series on Computational Intelligence (SSCI), 2016
The bi-objective just-in-time single-machine job-shop scheduling problem (JIT-JSP) aims at simult... more The bi-objective just-in-time single-machine job-shop scheduling problem (JIT-JSP) aims at simultaneously minimizing earliness and tardiness. In this paper, a multi-objective decoder-based evolutionary algorithm is proposed. The decoding strategy divides the search into two steps. In the first step, the search of the permutation order of the jobs is realized thanks to a multi-objective evolutionary algorithm. For a fixed permutation, the decoder algorithm optimizes the multi-objective timing sub-problem in the second step. Thus each permutation order induces a Pareto set of solutions. Two different decoding strategies to fix the idle times are proposed, one approximate and one exact. A comparison study with a classical multi-objective evolutionary algorithm underlines the performance of the proposed decoding strategy and the interest of the approximate decoder.
Journal of Heuristics, 2018
Metaheuristics are algorithms that have proven their efficiency on multi-objective combinatorial ... more Metaheuristics are algorithms that have proven their efficiency on multi-objective combinatorial optimisation problems. They often use local search techniques, either at their core or as intensification mechanisms, to obtain a well-converged and diversified final result. This paper surveys the use of local search techniques in multi-objective metaheuristics and proposes a general structure to describe and unify their underlying components. This structure can instantiate most of the multi-objective local search techniques and algorithms in literature.
Evolutionary Computation, 2018
Automatic algorithm configuration (AAC) is becoming a key ingredient in the design of high-perfor... more Automatic algorithm configuration (AAC) is becoming a key ingredient in the design of high-performance solvers for challenging optimisation problems. However, most existing work on AAC deals with configuration procedures that optimise a single performance metric of a given, single-objective algorithm. Of course, these configurators can also be used to optimise the performance of multi-objective algorithms, as measured by a single performance indicator. In this work, we demonstrate that better results can be obtained by using a native, multi-objective algorithm configuration procedure. Specifically, we compare three AAC approaches: one considering only the hypervolume indicator, a second optimising the weighted sum of hypervolume and spread, and a third that simultaneously optimises these complementary indicators, using a genuinely multi-objective approach. We assess these approaches by applying them to a highly-parametric local search framework for two widely studied multi-objective...
Ce travail présente nos principales contributions à la résolution de problèmes d'optimisation... more Ce travail présente nos principales contributions à la résolution de problèmes d'optimisation combinatoire en environnements déterministe et stochastique. Au niveau des métaheuristiques, une vue unifiée de la conception de métaheuristiques à solution unique et de métaheuristiques multi-objective est proposée. Cette unification a permis notamment de retravailler la plateforme ParadisEO afin d'offrir plus de flexibilité et de polyvalence. La synthèse des travaux présente également une vue unifiée des métaheuristiques coopératives. Nous montrons que cette vue convient aussi bien pour des coopérations entre métaheuristiques que des coopération entre des métaheuristiques et des méthodes exactes mais également des coopérations entre des métaheuristiques et des algorithmes d'extraction de connaissances. Différents exemples de coopérations réalisées dans mes travaux de recherche illustent ces coopérations et leur application à des problèmes d'optimisation combinatoire mono- ...
Lecture Notes in Computer Science, 2013
This paper focuses on the modeling and the implementation as a multi-objective optimization probl... more This paper focuses on the modeling and the implementation as a multi-objective optimization problem of a Pittsburgh classification rule mining algorithm adapted to large and imbalanced datasets, as encountered in hospital data. We associate to this algorithm an original post-processing method based on ROC curve to help the decision maker to choose the most interesting rules. After an introduction to problems brought by hospital data such as class imbalance, volumetry or inconsistency, we present MOCA-I-a Pittsburgh modelization adapted to this kind of problems. We propose its implementation as a dominance-based local search in opposition to existing multi-objective approaches based on genetic algorithms. Then we introduce the post-processing method to sort and filter the obtained classifiers. Our approach is compared to state-of-the-art classification rule mining algorithms, giving as good or better results, using less parameters. Then it is compared to C4.5 and C4.5-CS on hospital data with a larger set of attributes, giving the best results.
2009 IEEE Symposium on Computational Intelligence in Milti-Criteria Decision-Making, 2009
This paper gives a concise overview of evolutionary algorithms for multiobjective optimization. A... more This paper gives a concise overview of evolutionary algorithms for multiobjective optimization. A substantial number of evolutionary computation methods for multiobjective problem solving has been proposed so far, and an attempt of unifying existing approaches is here presented. Based on a fine-grained decomposition and following the main issues of fitness assignment, diversity preservation and elitism, a conceptual global model is proposed and is validated by regarding a number of state-of-the-art algorithms as simple variants of the same structure. The presented model is then incorporated into a generalpurpose software framework dedicated to the design and the implementation of evolutionary multiobjective optimization techniques: ParadisEO-MOEO. This package has proven its validity and flexibility by enabling the resolution of many real-world and hard multiobjective optimization problems.
Lecture Notes in Computer Science
This paper presents ParadisEO-MOEO, a white-box objectoriented generic framework dedicated to the... more This paper presents ParadisEO-MOEO, a white-box objectoriented generic framework dedicated to the flexible design of evolutionary multi-objective algorithms. This paradigm-free software embeds some features and techniques for Pareto-based resolution and aims to provide a set of classes allowing to ease and speed up the development of computationally efficient programs. It is based on a clear conceptual distinction between the solution methods and the multi-objective problems they are intended to solve. This separation confers a maximum design and code reuse. ParadisEO-MOEO provides a broad range of archive-related features (such as elitism or performance metrics) and the most common Pareto-based fitness assignment strategies (MOGA, NSGA, SPEA, IBEA and more). Furthermore, parallel and distributed models as well as hybridization mechanisms can be applied to an algorithm designed within ParadisEO-MOEO using the whole version of ParadisEO. In addition, GUIMOO, a platform-independant free software dedicated to results analysis for multi-objective problems, is briefly introduced.
Lecture Notes in Computer Science, 2015
In the single objective Unit Commitment Problem (UCP) the problem is usually separated in two sub... more In the single objective Unit Commitment Problem (UCP) the problem is usually separated in two sub-problems : the commitment problem which aims to fix the on/off scheduling of each unit and the dispatching problem which goal is to schedule the production of each turned on unit. The dispatching problem is a continuous convex problem that can easily be solved exactly. For the first sub-problem genetic algorithms (GA) are often applied and usually handle binary vectors representing the solutions of the commitment problem.Then the solutions are decoded in solving the dispatching problem with an exact method to obtain the precise production of each unit. In this paper a multiobjective version of the UCP taking the emission of gas into account is presented. In this multi objective UCP the dispatching problem remains easy to solve whereas considering it separatly remains interesting. A multi-objective GA handling binary vectors is applied. However for a binary representation there is a set of solutions of the dispatching problem that are pareto equivalent. Three decoding strategies are proposed and compared. The main contribution of this paper is the third decoding strategy which attaches an approximation of the Pareto front from the associated dispatching problem to each genotypic solution. It is shown that this decoding strategy leads to better results in comparison to the other ones.
Lecture Notes in Computer Science, 2008
The bi-objective ring star problem aims to locate a cycle through a subset of nodes of a graph wh... more The bi-objective ring star problem aims to locate a cycle through a subset of nodes of a graph while optimizing two types of cost. The first criterion is to minimize a ring cost, related to the length of the cycle, whereas the second one is to minimize an assignment cost, from non-visited nodes to visited ones. In spite of its natural multi-objective formulation, this problem has never been investigated in such a way. In this paper, three metaheuristics are designed to approximate the whole set of efficient solutions for the problem under consideration. Computational experiments are performed on well-known benchmark test instances, and the proposed methods are rigorously compared to each other using different performance metrics.
Revue d'intelligence artificielle, 2008
Bien que les algorithmes évolutionnaires soient couramment utilisés pour résoudre des problèmes m... more Bien que les algorithmes évolutionnaires soient couramment utilisés pour résoudre des problèmes multi-objectifs d'une part, et stochastiques d'autre part, très peu de travaux ont été menés sur ces deux aspects simultanément. Par exemple, les problèmes d'ordonnancement sont habituellement traités sous une forme mono-objectif déterministe, alors qu'ils sont clairement multi-objectifs et qu'ils sont soumis à de nombreux facteurs d'incertitude. Dans cet article, nous présentons différentes approches pour résoudre des problèmes d'optimisation multi-objectif stochastiques et les appliquons à un problème d'ordonnancement de type flow-shop de permutation bi-objectif avec durées d'exécution aléatoires. ABSTRACT. Although evolutionary algorithms are commonly used for solving multi-objective problems on the one hand and stochastic problems on the other hand, very few studies have investigated these two aspects simultaneously. For instance, scheduling problems are usually tackled in a single-objective deterministic form, whereas they are clearly multi-objective and they are subject to a wide range of uncertainty. In this paper, we present different approaches to solve stochastic multi-objective optimization problems and apply them to a bi-objective permutation flow-shop scheduling problem with random processing times.
Lecture Notes in Computer Science, 2013
The graph coloring problem is often investigated in the literature. Many insights about many neig... more The graph coloring problem is often investigated in the literature. Many insights about many neighboring solutions with the same fitness value are raised but as far as we know, no deep analysis of this neutrality has ever been conducted in the literature. In this paper, we quantify the neutrality of some hard instances of the graph coloring problem. This neutrality property has to be detected as it impacts the search process. Indeed, local optima may belong to plateaus that represents a barrier for local search methods. In this work, we also aim at pointing out the interest of exploiting neutrality during the search. Therefore, a generic local search dedicated to neutral problems, NILS, is performed on several hard instances.
Lecture Notes in Computer Science
This work deals with the design of new shielding materials for the protection of electrical devic... more This work deals with the design of new shielding materials for the protection of electrical devices. Since there are many different requirements for modern materials, we have chosen a multi-objective approach to this problem. As material under consideration we chose conducting polymer composites due to their excellent electromagnetic properties in the microwave band and their high potential for the optimization process. In this paper, we start this process with the formulation of a novel model, deal further with the approximation of these solution sets, and finally consider the decision support related to this problem.
Studies in Computational Intelligence, 2010
This chapter presents ParadisEO-MOEO, a white-box object-oriented software framework dedicated to... more This chapter presents ParadisEO-MOEO, a white-box object-oriented software framework dedicated to the flexible design of metaheuristics for multi-objective optimization. This paradigm-free software proposes a unified view for major evolutionary multi-objective metaheuristics. It embeds some features and techniques for multi-objective resolution and aims to provide a set of classes allowing to ease and speed up the development of computationally efficient programs. It is based on a clear conceptual distinction between the solution methods and the problems they are intended to solve. This separation confers a maximum design and code reuse. This general-purpose framework provides a broad range of fitness assignment strategies, the most common diversity preservation mechanisms, some elitistrelated features as well as statistical tools. Furthermore, a number of state-of-the-art search methods, including NSGA-II, SPEA2 and IBEA, have been implemented in a user-friendly way, based on the fine-grained ParadisEO-MOEO components.
Uploads
Papers by Laetitia Jourdan