An evolutionary algorithm recommendation method with a case study in flow shop scheduling

Abstract

Numerous optimization problems exist in the product and process design, engineering, and planning, as well as production management, and the evolutionary algorithms (EAs) have been proved to be effective optimization methods to solve these problems. However, how to choose the appropriate EA is one of the key issues due to the variety of EAs and lack of experience. Under this circumstance, firstly, a novel EA recommendation system framework is designed and proposed, and the implementation of the EA recommendation method is also described in this paper. Then, computational experiments are implemented to demonstrate the effectiveness and accuracy of the proposed method on the basis of 14 typical benchmark problems and 20 classical EAs. Finally, a case study regarding permutation flow shop scheduling is also carried out to test its effectiveness and accuracy. The proposed recommendation method provides a new way to select the appropriate EAs scientifically and reasonably to solve a given optimization problem.

Simulation of Smart Factory Processes Applying Multi-Agent-Systems—A Knowledge Management Perspective

“…Job scheduling and dynamic processing factors need to be considered in this context as well [29]. In addition, an improvement of tackling the problem of flow shop scheduling might be approached with the help of an evolutionary recommendation approach consisting of "collaborative filtering", a "ratings converter and some "recommendation assessment metrics [30].…”

EHHR: an efficient evolutionary hyper-heuristic based recommender framework for short-text classifier selection

“…In accordance with the “No Free Lunch theory [ 22 , 23 ],” there is no single heuristic that works best for all short-text classification problems. Brute force search of the best classifier for such big short-text data is expensive yet impractical in terms of time and computational resource utilization [ 24 ]. Additionally, selecting an acceptable heuristic for addressing a short-text classification problem is not simple since the problem’s features are typically not well understood in advance [ 25 ].…”