A Scalability Study and New Algorithms for Large-Scale Many-Objective Optimization
Keyword
Multi-objective OptimizationMany-objective Optimization
Computational Intelligence
Pareto Optimality
Optimization Algorithms
Metadata
Show full item recordAbstract
Many real-world optimization problems contain multiple (often conflicting) goals to be optimized concurrently, commonly referred to as multi-objective problems (MOPs). Over the past few decades, a plethora of multi-objective algorithms have been proposed, often tested on MOPs possessing two or three objectives. Unfortunately, when tasked with solving MOPs with four or more objectives, referred to as many-objective problems (MaOPs), a large majority of optimizers experience significant performance degradation. The downfall of these optimizers is that simultaneously maintaining a well-spread set of solutions along with appropriate selection pressure to converge becomes difficult as the number of objectives increase. This difficulty is further compounded for large-scale MaOPs, i.e., MaOPs possessing large amounts of decision variables. In this thesis, we explore the challenges of many-objective optimization and propose three new promising algorithms designed to efficiently solve MaOPs. Experimental results demonstrate the proposed optimizers to perform very well, often outperforming state-of-the-art many-objective algorithms.Collections
Related items
Showing items related by title, author, creator and subject.
-
Multi-Guide Particle Swarm Optimization for Large-Scale Multi-Objective Optimization ProblemsMulti-guide particle swarm optimization (MGPSO) is a novel metaheuristic for multi-objective optimization based on particle swarm optimization (PSO). MGPSO has been shown to be competitive when compared with other state-of-the-art multi-objective optimization algorithms for low-dimensional problems. However, to the best of the author’s knowledge, the suitability of MGPSO for high-dimensional multi-objective optimization problems has not been studied. One goal of this thesis is to provide a scalability study of MGPSO in order to evaluate its efficacy for high-dimensional multi-objective optimization problems. It is observed that while MGPSO has comparable performance to state-of-the-art multi-objective optimization algorithms, it experiences a performance drop with the increase in the problem dimensionality. Therefore, a main contribution of this work is a new scalable MGPSO-based algorithm, termed cooperative co-evolutionary multi-guide particle swarm optimization (CCMGPSO), that incorporates ideas from cooperative PSOs. A detailed empirical study on well-known benchmark problems comparing the proposed improved approach with various state-of-the-art multi-objective optimization algorithms is done. Results show that the proposed CCMGPSO is highly competitive for high-dimensional problems.
-
Comparative Study On Cooperative Particle Swarm Optimization Decomposition Methods for Large-scale OptimizationThe vast majority of real-world optimization problems can be put into the class of large-scale global optimization (LSOP). Over the past few years, an abundance of cooperative coevolutionary (CC) algorithms has been proposed to combat the challenges of LSOP’s. When CC algorithms attempt to address large scale problems, the effects of interconnected variables, known as variable dependencies, causes extreme performance degradation. Literature has extensively reviewed approaches to decomposing problems with variable dependencies connected during optimization, many times with a wide range of base optimizers used. In this thesis, we use the cooperative particle swarm optimization (CPSO) algorithm as the base optimizer and perform an extensive scalability study with a range of decomposition methods to determine ideal divide-and-conquer approaches when using a CPSO. Experimental results demonstrate that a variety of dynamic regrouping of variables, seen in the merging CPSO (MCPSO) and decomposition CPSO (DCPSO), as well varying total fitness evaluations per dimension, resulted in high-quality solutions when compared to six state-of-the-art decomposition approaches.
-
Implementing the OPTIMAL model : the impact on students' motivation in an elementary school games environmentOptimal challenge occurs when an individual perceives the challenge of the task to be equaled or matched by his or her own skill level (Csikszentmihalyi, 1990). The purpose of this study was to test the impact of the OPTIMAL model on physical education students' motivation and perceptions of optimal challenge across four games categories (i. e. target, batting/fielding, net/wall, invasion). Enjoyment, competence, student goal orientation and activity level were examined in relation to the OPTIMAL model. A total of 22 (17 M; 5 F) students and their parents provided informed consent to take part in the study and were taught four OPTIMAL lessons and four non-OPTIMAL lessons ranging across the four different games categories by their own teacher. All students completed the Task and Ego in Sport Questionnaire (TEOSQ; Duda & Whitehead, 1998), the Intrinsic Motivation Inventory (IMI; McAuley, Duncan, & Tanmien, 1987) and the Children's Perception of Optimal Challenge Instrument (CPOCI; Mandigo, 2001). Sixteen students (two each lesson) were observed by using the System for Observing Fitness Instruction Time tool (SOFTT; McKenzie, 2002). As well, they participated in a structured interview which took place after each lesson was completed. Quantitative results concluded that no overall significant difference was found in motivational outcomes when comparing OPTIMAL and non-OPTIMAL lessons. However, when the lessons were broken down into games categories, significant differences emerged. Levels of perceived competence were found to be higher in non-OPTIMAL batting/fielding lessons compared to OPTIMAL lessons, whereas levels of enjoyment and perceived competence were found to be higher in OPTIMAL invasion lessons in comparison to non-OPTIMAL invasion lessons. Qualitative results revealed significance in feehngs of skill/challenge balance, enjoyment and competence in the OPTIMAL lessons. Moreover, a significance of practically twice the active movement time percentage was found in OPTIMAL lessons in comparison to non-OPTIMAL lessons.
