Hybrid metaheuristic approach for handling many objectives and decisions on partial support in project portfolio optimisation

Organisations often face portfolio optimisation problems. In many practical cases, the decision is not only to select a subset of applicant projects but also to assign a number of resources to the favoured proposals within the projects' feasibilities. In the scientific literature, multi-objective optimisation algorithms have addressed this issue by means of generating redundant dummy projects for each proposal that is likely to be partially funded. However, unfortunately, if there are many projects that can be supported in a wide variety of ways, this approximation results in a large overload for the optimisation methods, which provokes poor algorithmic performance. To alleviate these problems, we propose the Non-Outranked Ant Colony Optimisation II method, which incorporates a fuzzy outranking preference model for optimising portfolio problems with partial-support features. The advantages offered by this novel approach are supported by a series of experiments that provide evidence of its capacity for solving those real-world problems in which the level of resources allocated to the selected proposals has a proportional impact on the projects' expected benefits.

[1]  Peter J. Fleming,et al.  On the Evolutionary Optimization of Many Conflicting Objectives , 2007, IEEE Transactions on Evolutionary Computation.

[2]  Vincent Mousseau,et al.  Inferring an ELECTRE TRI Model from Assignment Examples , 1998, J. Glob. Optim..

[3]  Jindrich Klapka,et al.  Multicriterial Projects Selection , 2013, Handbook of Optimization.

[4]  S. Ghorbani,et al.  A new multi-objective algorithm for a project selection problem , 2009, Adv. Eng. Softw..

[5]  Richard F. Hartl,et al.  Pareto ant colony optimization with ILP preprocessing in multiobjective project portfolio selection , 2006, Eur. J. Oper. Res..

[6]  S French,et al.  Multicriteria Methodology for Decision Aiding , 1996 .

[7]  Chiuh-Cheng Chyu,et al.  Applying Memetic Algorithm in Multi-Objective Resource Allocation among Competing Projects , 2010, J. Softw..

[8]  Jean Pierre Brans,et al.  HOW TO SELECT AND HOW TO RANK PROJECTS: THE PROMETHEE METHOD , 1986 .

[9]  Mitsuo Gen,et al.  Multi-criteria human resource allocation for solving multistage combinatorial optimization problems using multiobjective hybrid genetic algorithm , 2008, Expert Syst. Appl..

[10]  Jaap Spronk,et al.  Erim Report Series Research in Management a Framework for Managing a Portfolio of Socially Responsible Investments Bibliographic Data and Classifications , 2002 .

[11]  Rafael Caballero,et al.  Solving a comprehensive model for multiobjective project portfolio selection , 2010, Comput. Oper. Res..

[12]  Günther R. Raidl,et al.  An improved hybrid genetic algorithm for the generalized assignment problem , 2004, SAC '04.

[13]  Stefan Biffl,et al.  Applied Soft Computing Software Project Portfolio Optimization with Advanced Multiobjective Evolutionary Algorithms , 2022 .

[14]  Norman P. Archer,et al.  Project portfolio selection through decision support , 2000, Decis. Support Syst..

[15]  Juan Gaytán Iniestra,et al.  Multicriteria decisions on interdependent infrastructure transportation projects using an evolutionary-based framework , 2009, Appl. Soft Comput..

[16]  Babak Amiri,et al.  A Multi-Objective Hybrid Optimization Algorithm for Project Selection Problem , 2012 .

[17]  Bernard Roy.  Actions and Decision Aiding , 1996 .

[18]  Ahti Salo,et al.  Robust portfolio modeling with incomplete cost information and project interdependencies , 2008, Eur. J. Oper. Res..

[19]  Matthew O. Ward,et al.  Hierarchical parallel coordinates for exploration of large datasets , 1999, Proceedings Visualization '99 (Cat. No.99CB37067).

[20]  Yannis Siskos,et al.  Preference disaggregation: 20 years of MCDA experience , 2001, Eur. J. Oper. Res..

[21]  Eduardo Fernández,et al.  Large-Scale Public R&D Portfolio Selection by Maximizing a Biobjective Impact Measure , 2010, IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans.

[22]  Andrés L. Medaglia,et al.  A multiobjective evolutionary approach for linearly constrained project selection under uncertainty , 2007, Eur. J. Oper. Res..

[23]  E. Wegman.  Hyperdimensional Data Analysis Using Parallel Coordinates , 1990 .

[24]  E. Fernandez,et al.  Computer-based decision models for R&D project selection in public organizations , 2005 .

[25]  G. A. Miller.  THE PSYCHOLOGICAL REVIEW THE MAGICAL NUMBER SEVEN, PLUS OR MINUS TWO: SOME LIMITS ON OUR CAPACITY FOR PROCESSING INFORMATION 1 , 1956 .

[26]  Peter J. Fleming,et al.  Many-Objective Optimization: An Engineering Design Perspective , 2005, EMO.

[27]  G. Montibeller,et al.  Resource Allocation in Local Government with Facilitated Portfolio Decision Analysis , 2011 .

[28]  Igor Litvinchev,et al.  A milp bi-objective model for static portfolio selection of R&D projects with synergies , 2011 .

[29]  Martin Rogers,et al.  Choosing realistic values of indifference, preference and veto thresholds for use with environmental criteria within ELECTRE , 1998, Eur. J. Oper. Res..

[30]  Fereidoun Ghasemzadeh,et al.  A zero-one model for project portfolio selection and scheduling , 1999, J. Oper. Res. Soc..

[31]  George M. Marakas,et al.  Decision support systems in the twenty-first century , 1998 .

[32]  Luis C. Dias,et al.  Inferring Electre's veto-related parameters from outranking examples , 2006, Eur. J. Oper. Res..

[33]  Ana F. Carazo,et al.  A project portfolio selection problem in a group decision-making context , 2011 .

[34]  Eduardo Fernández,et al.  Many-Objective Portfolio Optimization of Interdependent Projects with 'a priori' Incorporation of Decision-Maker Preferences , 2014 .

[35]  Jie Yang,et al.  A Dynamic and Fuzzy Modeling Approach for Multi-objective R&D Project Portfolio Selection , 2012 .

[36]  Günther R. Raidl,et al.  Combining Metaheuristics and Exact Algorithms in Combinatorial Optimization: A Survey and Classification , 2005, IWINAC.

[37]  Jindrich Klapka,et al.  Decision support system for multicriterial R&D and information systems projects selection , 2002, Eur. J. Oper. Res..

[38]  Bernard Roy,et al.  Determining the weights of criteria in the ELECTRE type methods with a revised Simos' procedure , 2002, Eur. J. Oper. Res..

[39]  José Rui Figueira,et al.  Using assignment examples to infer weights for ELECTRE TRI method: Some experimental results , 2001, Eur. J. Oper. Res..

[40]  Carlos A. Coello Coello,et al.  Preference incorporation to solve many-objective airfoil design problems , 2011, 2011 IEEE Congress of Evolutionary Computation (CEC).

[41]  Magdalene Marinaki,et al.  An evolutionary approach to construction of outranking models for multicriteria classification: The case of the ELECTRE TRI method , 2009, Eur. J. Oper. Res..

[42]  Eduardo Fernández,et al.  Application of the non-outranked sorting genetic algorithm to public project portfolio selection , 2013, Inf. Sci..

[43]  Agnès Plateau,et al.  A hybrid search combining interior point methods and metaheuristics for 0–1 programming , 2002 .

[44]  Christian Stummer,et al.  Interactive R&D portfolio analysis with project interdependencies and time profiles of multiple objectives , 2003, IEEE Trans. Engineering Management.

[45]  Konstantinos Liagkouras,et al.  Multiobjective Evolutionary Algorithms for Portfolio Management: A comprehensive literature review , 2012, Expert Syst. Appl..

[46]  Eduardo Fernández,et al.  Increasing selective pressure towards the best compromise in evolutionary multiobjective optimization: The extended NOSGA method , 2011, Inf. Sci..

[47]  El-Ghazali Talbi,et al.  Hybridizing exact methods and metaheuristics: A taxonomy , 2009, Eur. J. Oper. Res..

[48]  Richard F. Hartl,et al.  Pareto Ant Colony Optimization: A Metaheuristic Approach to Multiobjective Portfolio Selection , 2004, Ann. Oper. Res..

[49]  B. Roy.  THE OUTRANKING APPROACH AND THE FOUNDATIONS OF ELECTRE METHODS , 1991 .