Addressing expensive multi-objective games with postponed preference articulation via memetic co-evolution

Żychowski, Adam; Gupta, Abhishek; Mańdziuk, Jacek; Ong, Yew-Soon

doi:10.1016/j.knosys.2018.05.012

Cited by 12 publications

(3 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Various perspectives have recently been put forward on how the multiple solutions returned by the lower level are to be handled at the upper level. This includes a so-called optimistic formulation in [81,82], as well as a pessimistic / adversarial / worst-case formulation in [83,84]. In either case, the overall multi-objective bi-level optimization problem is shown to offer an ideal setting for the implicit parallelism of EAs to be harnessed.…”

Section: Outline Of Methodologiesmentioning

confidence: 99%

Back to the Roots: Multi-X Evolutionary Computation

Gupta

Ong

2019

Cogn Comput

Self Cite

View full text Add to dashboard Cite

BACKGROUND: Over the years, evolutionary computation has come to be recognized as one of the leading algorithmic paradigms in the arena of global black-box optimization. The distinguishing facets of evolutionary methods, inspired by Darwin's foundational principles of natural selection, stem mainly from their population-based search strategywhich gives rise to the phenomenon of implicit parallelism. Precisely, even as an evolutionary algorithm manipulates a population of a few candidate solutions (or: individuals), it is able to simultaneously sample, evaluate, and process a vast number of regions of the search space. This behavior is in effect analogous to our inherent cognitive ability of processing diverse information streams (such as sight and sound) with apparent simultaneity in different regions of our brain. METHODS:For this reason, evolutionary algorithms have emerged as the method of choice for those search and optimization problems where a collection of multiple target solutions (that may be scattered throughout the search space) are to be found in a single run. With the above in mind, in this paper we return to the roots of evolutionary computation, with the aim of shedding light on a variety of problem settings that are uniquely suited for exploiting the implicit parallelism of evolutionary algorithms. RESULTS:Our discussions cover established concepts of multi-objective and multi-modal optimization, as well as new (schema) theories pertaining to emerging problem formulations that entail multiple searches to be carried out at once. We capture associated research activities under the umbrella term of multi-X evolutionary computation, where X, as of now, represents the following list: {'objective ', 'modal', 'task', 'level', 'hard', 'disciplinary', 'form'}. CONCLUSIONS: With this, we hope that the present position paper will serve as a catalyst for effecting further research efforts into such areas of optimization problem-solving that are well-aligned with the fundamentals of evolutionary computation; in turn prompting the steady update of the list X with new applications in the future.

show abstract

Section: Outline Of Methodologiesmentioning

confidence: 99%

Back to the Roots: Multi-X Evolutionary Computation

Gupta

Ong

2019

Cogn Comput

Self Cite

View full text Add to dashboard Cite

show abstract

“…Second, artificial intelligence-based approaches are used to solve the multicriteria game. For example, the fuzzy set-based algorithm can calculate PNE solutions by constructing membership functions for fuzzy goals [28,29], and the heuristic algorithm can solve it by using a population evolution strategy [30,31]. However, these algorithms are usually used in problems with low real-time requirements, and their application range is always limited for the problem constraints.…”

Section: Introductionmentioning

confidence: 99%

Multicriteria Game Approach to Air-to-Air Combat Tactical Decisions for Multiple UAVs

Jiang,

Luo,

et al. 2023

J. of Syst. Eng. Electron.

View full text Add to dashboard Cite

Air-to-air combat tactical decisions for multiple unmanned aerial vehicles (ACTDMU) are a key decision-making step in beyond visual range combat. Complex influencing factors, strong antagonism and real-time requirements need to be considered in the ACTDMU problem. In this paper, we propose a multicriteria game approach to ACTDMU. This approach consists of a multicriteria game model and a Pareto Nash equilibrium algorithm. In this model, we form the strategy profiles for the integration of air-to-air combat tactics and weapon target assignment strategies by considering the correlation between them, and we design the vector payoff functions based on predominance factors. We propose a algorithm of Pareto Nash equilibrium based on preference relations using threshold constraints (PNE-PRTC), and we prove that the solutions obtained by this algorithm are refinements of Pareto Nash equilibrium solutions. The numerical experiments indicate that PNE-PRTC algorithm is considerably faster than the baseline algorithms and the performance is better. Especially on large-scale instances, the Pareto Nash equilibrium solutions can be calculated by PNE-PRTC algorithm at the second level. The simulation experiments show that the multicriteria game approach is more effective than one-side decision approaches such as multiple-attribute decision-making and randomly chosen decisions.

show abstract

“…In _ Zychowski et al (2018), a memetic co-evolution approach is presented which aims at overcoming the Red Queen effect when solving MOGs. Moshaiov, 2018).…”

mentioning

confidence: 99%

Solving zero‐sum multi‐objective games with a‐priori secondary criteria

Harel

Eisenstadt-Matalon

Moshaiov

2022

Multi Criteria Decision Anal

View full text Add to dashboard Cite

Solving non‐cooperative zero‐sum multi‐objective Games (zsMOGs), under undecided objective preferences results, for each of the players, in a Set of Rationalizable Strategies (SRS) to choose from. First, this paper deals with finding for each of the players a preferred subset of such rationalizable strategies based on a‐priori incorporation of partial preferences of the decision‐makers using secondary criteria. The obtained subset is termed the Set of Preferred Strategies (SPS). Here, a novel archive‐based co‐evolutionary algorithm is suggested to search for the SPS for each of the players. An academic example is suggested to demonstrate and validate the algorithm. It concerns a zsMOG that involves two adversarial planar manipulators. Based on a theorem that is proven here, a theoretic reference SRS is found for each of the players. This reference SRS is applied to find a reference SPS, which is used for validating the algorithm. Next, a comparison study is performed between the proposed archive‐based co‐evolutionary algorithm and an elite‐based version of this algorithm. The results clearly show that the archive‐based algorithm is superior to the elite‐based version, yielding results that correspond well to the theoretic sets.

show abstract

Addressing expensive multi-objective games with postponed preference articulation via memetic co-evolution

Cited by 12 publications

References 23 publications

Back to the Roots: Multi-X Evolutionary Computation

Back to the Roots: Multi-X Evolutionary Computation

Multicriteria Game Approach to Air-to-Air Combat Tactical Decisions for Multiple UAVs

Solving zero‐sum multi‐objective games with a‐priori secondary criteria

Contact Info

Product

Resources

About