Evolutionary Multitask Optimization: a Methodological Overview, Challenges and Future Research Directions

Abstract

In this work we consider multitasking in the context of solving multiple optimization problems simultaneously by conducting a single search process. The principal goal when dealing with this scenario is to dynamically exploit the existing complementarities among the problems (tasks) being optimized, helping each other through the exchange of valuable knowledge. Additionally, the emerging paradigm of Evolutionary Multitasking tackles multitask optimization scenarios by using as inspiration concepts drawn from Evolutionary Computation. The main purpose of this survey is to collect, organize and critically examine the abundant literature published so far in Evolutionary Multitasking, with an emphasis on the methodological patterns followed when designing new algorithmic proposals in this area (namely, multifactorial optimization and multipopulation-based multitasking). We complement our critical analysis with an identification of challenges that remain open to date, along with promising research directions that can stimulate future efforts in this topic. Our discussions held throughout this manuscript are offered to the audience as a reference of the general trajectory followed by the community working in this field in recent times, as well as a self-contained entry point for newcomers and researchers interested to join this exciting research avenue.

Jack and Masters of All Trades: One-Pass Learning of a Set of Model Sets from Foundation Models

“…( 5) in a principled manner; it is simple to implement without requiring ad hoc transfer mechanisms [46]. Detailed reviews on alternative algorithmic techniques that can be considered are provided in [55,56].…”

Evolutionary Multitask Optimization: Fundamental Research Questions, Practices, and Directions for the Future

“…When solving real-world optimization problems, do we really obtain a profit by addressing them together via multitasking when compared to the case when the problems are solved in isolation from each other with competitive single-task optimization algorithms? After years of activity that have been summarized in recent surveys on Evolutionary Multitasking [7,8], we firmly believe that it is the moment to expose and reflect these crucial concerns. Solid and informed answers to these fundamental questions are still lacking, which can lead to undesirable developments and outcomes of no practical value in the future of this field.…”

MO-MFCGA: Multiobjective Multifactorial Cellular Genetic Algorithm for Evolutionary Multitasking

“…The principal motivation for the adoption of concepts from evolutionary computation in this area is twofold: i) the inherent parallelism granted by a population of solutions, providing an suitable framework for dealing with concurrent tasks, and ii) the maintenance of a set of solutions over the search, which eases the exploration of synergies among the tasks and the exchange of genetic material among individuals [12]. Diverse perspectives have been given over the years for formalizing the EM concept, reviewed recently in comprehensive surveys on the matter [7]. However, there is a clear consensus in the literature around the capital role of multifactorial optimization as a leader paradigm to realize EM [15], with MFEA [12] as one of the most popular algorithms resulting from this paradigm.…”

“…Also valuable is the study in [19], in which MFEA is used to solve several mobile agents path planning problems at once. Further related contributions can be found in the literature review offered in [7].…”

“…Although MOPs have been so far tackled via different solvers, it has not been until recently when Transfer Optimization [6] has emerged as a promising paradigm for this kind of optimization problems. Transfer Optimization is an incipient research field with a growing momentum in the current research panorama [7]. The main inspiration underneath this paradigm is the exploitation of the knowledge generated during the optimization of one problem for efficiently solving another related or unrelated problem.…”