Machine Learning for Enhancing Metaheuristics in Global Optimization: A Comprehensive Review

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper presents a comprehensive review of the integration of machine learning (ML) techniques with metaheuristic optimization for global optimization problems. The authors identify the inherent limitations of traditional metaheuristics—such as parameter sensitivity, premature convergence, and scalability issues—and state that ML offers a powerful, data-driven paradigm to address these challenges.
Here are my concerns
1. Section 5.4 is very current and interesting. However, it could be slightly expanded to explain why these models are particularly well-suited for offline optimization compared to other generative models like GANs, beyond the brief mention of avoiding mode collapse. 
2. There appear to be some duplicate citations. For example, [69] and [28] are the same paper.
I have no more comments since this paper is well-structured and well-written. I think this paper can be accepted after minor revisions.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study reviews how machine learning techniques can improve metaheuristic optimization algorithms. That is, it reviews integration approaches, analyzes representative methods, and highlights emerging trends like learned optimizers, explainable surrogates, and diffusion models. I found the paper well-written and structured. Therefore, I should thank the authors for writing such an interesting review. However, the paper can be improved by considering the following details:

Although machine learning techniques can improve metaheuristic optimization algorithms, metaheuristic optimization algorithms have been widely used to enhance machine learning models, such as tuning their hyperparameters, optimizing the data split ratio (how best to split the training and testing data), and feature selection. Hence, adding a section to shortly discuss these methods can increase the value of the paper.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This document is a review of the application of machine learning enhanced meta-heuristic algorithms in global optimization, introducing the limitations of premature convergence and parameter sensitivity of meta-heuristic algorithms, as well as the role of machine learning in solving these problems, and proposes a classification system based on the functional role of machine learning in the optimization process, covering categories such as learning or adaptive search operators, proxy modeling, etc., and also discusses the application of related machine learning technologies, emerging trends, Open up challenges and future directions. Here are some of my questions:

1.This paper proposes "green optimization" in the future direction, but does not specify how to quantify the energy consumption cost of the optimization algorithm.

2.Regarding benchmarking, the article notes that ML-enhanced algorithms have "reproducibility issues" but does not mention specific solutions.

3.Regarding the meta-learning part, the article emphasizes its "cross-task migration" capabilities, but most cases rely on large-scale training data. For example, for the practical problem of data scarcity, what do the authors think is the practical value of meta-learning methods?

4.Regarding the classification system of machine learning and meta-heuristic algorithms, a taxonomy based on functional roles is proposed, but there may be cross-coupling between different categories. How do authors define the core boundaries of each category?

5. In this paper, authors focus on machine learning for global optimization. THE importance of AI optimization schemes for application need to be compared and analyzed to reflect the advantages of your work, which can refer to：

[a] IEEE Transactions on Industrial Informatics, DOI: 10.1109/TII.2024.3390595

[b] CSEE Journal of Power and Energy Systems, vol. 8, no. 1, pp. 95-104, Jan. 2022

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

see attachment

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

N