Achieving Highly Scalable Evolutionary Real-Valued Optimization by Exploiting Partial Evaluations

Bouter, Anton; Alderliesten, Tanja; Bosman, Peter A. N.

doi:10.1162/evco_a_00275

Cited by 10 publications

(4 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Bouter et al, 2021: In [ 92 ], the authors explore the possibilities of accelerating the speed of an evolutionary algorithm applied to a DIR problem. The speed advantage is gained by profiting on the parallelization potential of the Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) [ 93 ] and its adaptation for Real Value genes RV-GOMEA [ 94 ]. Evaluation and evolution are performed in parallel groups of individuals that are independent: no individual in a group is correlated to an individual in another group.…”

Section: Articlesmentioning

confidence: 99%

Evolutionary Image Registration: A Review

Cocianu

Uscatu²,

Stan³

2023

Sensors

View full text Add to dashboard Cite

Image registration is one of the most important image processing tools enabling recognition, classification, detection and other analysis tasks. Registration methods are used to solve a large variety of real-world problems, including remote sensing, computer vision, geophysics, medical image analysis, surveillance, and so on. In the last few years, nature-inspired algorithms and metaheuristics have been successfully used to address the image registration problem, becoming a solid alternative for direct optimization methods. The aim of this paper is to investigate and summarize a series of state-of-the-art works reporting evolutionary-based registration methods. The papers were selected using the PRISMA 2020 method. The reported algorithms are reviewed and compared in terms of evolutionary components, fitness function, image similarity measures and algorithm accuracy indexes used in the alignment process.

show abstract

Section: Articlesmentioning

confidence: 99%

Evolutionary Image Registration: A Review

Cocianu

Uscatu²,

Stan³

2023

Sensors

View full text Add to dashboard Cite

show abstract

“…This is the prominent area of research where, based on a deeper understanding of the problem, efficient operators are possible to design with significant improvement in in computational time, search efficiency, or both. A large number of works essentially belonging to gray-box optimization implement mutation operators by essentially synthesizing the difference vectors and storing only a few complete individuals [5,6,17]. These works either use only mutation operators or invoke crossover operators only infrequently to amortize their cost.…”

Section: Preliminaries and Related Workmentioning

confidence: 99%

Improving Time and Memory Efficiency of Genetic Algorithms by Storing Populations as Minimum Spanning Trees of Patches

Buzdalov

2023

Proceedings of the Companion Conference on Genetic and Evolutionary Computation

View full text Add to dashboard Cite

In many applications of evolutionary algorithms the computational cost of applying operators and storing populations is comparable to the cost of fitness evaluation. Furthermore, by knowing what exactly has changed in an individual by an operator, it is possible to recompute fitness value much more efficiently than from scratch. The associated time and memory improvements have been available for simple evolutionary algorithms, few specific genetic algorithms and in the context of gray-box optimization, but not for all algorithms, and the main reason is that it is difficult to achieve in algorithms using large arbitrarily structured populations.This paper makes a first step towards improving this situation. We show that storing the population as a minimum spanning tree, where vertices correspond to individuals but only contain meta-information about them, and edges store structural differences, or patches, between the individuals, is a viable alternative to the straightforward implementation. Our experiments suggest that significant, even asymptotic, improvements -including execution of crossover operators! -can be achieved in terms of both memory usage and computational costs.

show abstract

“…To prevent having to tune the population size of each of the algorithms by hand, we use the Interleaved Multi-start Scheme (IMS) [6].…”

Section: Interleaved Multi-start Schemementioning

confidence: 99%

“…problem decompositions [11], or partial evaluations [6], to achieve great success where other optimization methods failed. In this paper, we specifically consider a GBO setting where partial evaluations, by which we mean that evaluating the change in fitness after only a few variables have changed, can be done (proportionally) more efficiently than when a full evaluation is performed.…”

Section: Introductionmentioning

confidence: 99%

GPU-Accelerated Parallel Gene-pool Optimal Mixing in a Gray-Box Optimization Setting

Bouter¹,

Bosman²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

In a Gray-Box Optimization (GBO) setting that allows for partial evaluations, the fitness of an individual can be updated efficiently after a subset of its variables has been modified. This enables more efficient evolutionary optimization with the Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) due to its key strength: Gene-pool Optimal Mixing (GOM). For each solution, GOM performs variation for many (small) sets of variables. To improve efficiency even further, parallel computing can be leveraged. For EAs, typically, this comprises population-wise parallelization. However, unless population sizes are large, this offers limited gains. For large GBO problems, parallelizing GOM-based variation holds greater speed-up potential, regardless of population size. However, this potential cannot be directly exploited because of dependencies between variables. We show how graph coloring can be used to group sets of variables that can undergo variation in parallel without violating dependencies. We test the performance of a CUDA implementation of parallel GOM on a Graphics Processing Unit (GPU) for the Max-Cut problem, a well-known problem for which the dependency structure can be controlled. We find that, for sufficiently large graphs with limited connectivity, finding high-quality solutions can be achieved up to 100 times faster, showcasing the great potential of our approach. CCS CONCEPTS• Mathematics of computing → Evolutionary algorithms.

show abstract

Achieving Highly Scalable Evolutionary Real-Valued Optimization by Exploiting Partial Evaluations

Cited by 10 publications

References 50 publications

Evolutionary Image Registration: A Review

Evolutionary Image Registration: A Review

Improving Time and Memory Efficiency of Genetic Algorithms by Storing Populations as Minimum Spanning Trees of Patches

GPU-Accelerated Parallel Gene-pool Optimal Mixing in a Gray-Box Optimization Setting

Contact Info

Product

Resources

About