Meta-Learning for Optimization: A Case Study on the Flowshop Problem Using Decision Trees

Pavelski, Lucas Marcondes; Delgado, Myriam Regattieri; Kessaci, Marie‐Éléonore

doi:10.1109/cec.2018.8477664

Cited by 7 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The proposal and results presented in this paper extend previous works developed by Pavelski et al (2018aPavelski et al ( , 2018bPavelski et al ( , 2019. As far as we know, no other PIAC proposal in the literature uses AAC and FLA to build the meta-data.…”

Section: Introductionsupporting

confidence: 89%

Stochastic local search and parameters recommendation: a case study on flowshop problems

Pavelski

Delgado

Kessaci

et al. 2020

Int Trans Operational Res

Self Cite

View full text Add to dashboard Cite

The Algorithm Selection Problem (ASP) considers the use of previous knowledge regarding problem features and algorithm performance to recommend the best strategy to solve a previously unseen problem. In the application context, the usual ASP for optimization considers recommending the best heuristics, whenever it faces a new similar problem instance, also known as the Per-Instance ASP. Although ASP for heuristic recommendation is not new, selecting heuristics and also their parameters, or the Per-instance Algorithm Configuration Problem, is still considered a challenging task. This paper investigates the use of meta-learning to recommend six different stochastic local searches and their parameters to solve several instances of permutation flowshop problems. The proposed approach uses several problem features, including fitness landscape metrics, builds the performance database using irace, and trains different multi-label recommendation models on a data set with more than 6000 flowshop problem instances. Experiments show that decision tree-based machine learning models achieve good performance, and the quality of the recommendations is capable of outperforming the state-of-the-art algorithm with tuned configuration.

show abstract

Section: Introductionsupporting

confidence: 89%

Stochastic local search and parameters recommendation: a case study on flowshop problems

Pavelski

Delgado

Kessaci

et al. 2020

Int Trans Operational Res

Self Cite

View full text Add to dashboard Cite

show abstract

“…Instance dissimilarity and algorithmic discrimination are two crucial ASP criteria. The algorithms require a variety of instances in order to provide such a rich collection of data [96]. If there are many instances produced, it could take a long time to execute all candidate algorithms on them.…”

Section: Discussionmentioning

confidence: 99%

Learnheuristics in routing and scheduling problems: A review

Hussein¹,

Yaseen²,

Rashid³

2023

sjpas

View full text Add to dashboard Cite

Combinatorial optimization problems (COPs) are the most important class of optimization problems, with great practical significance. This class is concerned with identifying the best solution from a discrete set of all available options. The transportation (routing) and distribution (scheduling) systems are considered the most challenging optimization examples of the COPs. Given the importance of routing and scheduling problems, many methods have been proposed to address them. These methods can be categorized into traditional (exact and metaheuristics (MHs) methods) and machine learning (ML) methods. ML methods have been proposed to overcome the problems that traditional methods suffer from, especially high computational time and dependence on the knowledge of experts. Recently, ML methods and MHs have been combined to tackle the COPs, and then the learnheuristics term emerged. This combination aims to guide the MHs toward an efficient, effective, and robust search and improve their performance in terms of solution quality. This work reviews the publications in which the collaboration between MHs and ML has been utilized to propose a guideline for the researchers to put forward new algorithms that have a good ability to solve routing and scheduling problems.

show abstract

Fundamental Machine Learning Methods

2024

Machine Learning and Metaheuristic Computation

View full text Add to dashboard Cite

Meta-Learning for Optimization: A Case Study on the Flowshop Problem Using Decision Trees

Cited by 7 publications

References 17 publications

Stochastic local search and parameters recommendation: a case study on flowshop problems

Stochastic local search and parameters recommendation: a case study on flowshop problems

Learnheuristics in routing and scheduling problems: A review

Fundamental Machine Learning Methods

Contact Info

Product

Resources

About