Dynamic Population on Bio-Inspired Algorithms Using Machine Learning for Global Optimization

Round 1

Reviewer 1 Report

The authors of the manuscript number 2746951 propose a modified optimization algorithm coupling multiple approaches and adapting the population size dynamically. The population size adaptation is appealing and relevant to the scientific community. The article should be however improved and corrected before publication. In what follows you find some remarks to address before final publication:

·      Page 5 line 41 equation (1) is not introduced and doesn’t really fit in its place in the text.

·      Page 6 line 159: this enumeration is out of place.

·      Algorithm 4 is unclear: What are your Diversification criteria and Intensification criteria, how do you rank your solution? How does DBSCAN results affect your algorithm?

·      Page 10 line 227: retrieved from the.  (the literature?)

·      Can you list the equations of the used functions in table 5

·      The comparison is well performed in terms of accuracy; however, the computation time is not shown. The reviewer believes that the computation is significantly affected by the dynamic increase of the population size. The authors alleviate this effect by setting an upper limit of the population size as implied in figures 21 and 23. The reviewer believes the computation time comparison is important to confirm/reject the superiority of the proposed algorithm.

Some typos and missing text should be corrected.

Author Response

Dear Reviewer,

Thank you for your valuable time and insightful feedback on our manuscript. Your observations have been carefully addressed and are attached in the accompanying PDF. We look forward to receiving further comments if there are any concerns.

Best Regards,

Nicolás Caselli B.

Author Response File: Author Response.pdf

Reviewer 2 Report

1. The description of the cuckoo method is not accurate. It is not clear why the Hadamard product is used in (1), since the formula is written in coordinate form. The given pseudocode does not fully correspond to the algorithm outlined in 46.

2. When describing the PSO method, not a single formula is given (for the velocity and position of the particle).

3.Sections 3.2 and 3.3 do not contain a constructive description of the proposed actions.

4. In general, the article needs an explanation of the novelty of the proposed approach and its detailed description. The ideas of using clustering and dynamically changing the population size in metaheuristic optimization algorithms are not new.

5. Obtained in Sect. 4.2 The numerical results require additional explanations due to the potential novelty of the work.

Author Response

Dear Reviewer,

Thank you for your valuable time and insightful feedback on our manuscript. Your observations have been carefully addressed and are attached in the accompanying PDF. Your contributions have significantly improved the quality of our work.

Best Regards,

Nicolás Caselli Benavente.

Author Response File: Author Response.pdf

Reviewer 3 Report

"Metaheuristic algorithms, especially their autonomous variants, are emerging as promising tools to address this challenge."

What the authora mean by "auonomous", and  why "especially"?

In (1), the element d of a solution i at iteration t should probably bw denoted as x_i^d(t).

No puncryation marks in formulas.

The level of detalization of the algorithms in Section 3.1 differs very much. E.g., from Algorithm 1, it is impossible to reconstruct the whole Cuckoo algorithm.

The same for Alg. 3,2.

Author Response

Dear Reviewer,

Thank you for your valuable time and insightful feedback on our manuscript. Your observations have been carefully addressed and are attached in the accompanying PDF. Your contributions have significantly improved the quality of our work.

Best Regards,

Nicolás Caselli Benavente.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thank you for the corrections and the clarifications

Good

Author Response

Dear reviewer,

We would like to express our gratitude for your time and valuable insights that have helped improve our research.

We have included this acknowledgment in the attached PDF document. 

Best regards!

Author Response File: Author Response.pdf

Reviewer 2 Report

1. Equation (1) describes the change in one coordinate of the solution vector, so the use of the Hadamard vector multiplication operation does not make sense in this context.

2. The reviewer's second comment has been taken into account.

3.4. The third and fourth comments are not taken into account sufficiently. The authors chose a linguistic way to describe the result. According to the reviewer, a more rigorous and constructive mathematical description is required. The goal is to achieve repeatability of the results obtained by potential readers.

5. The fifth comment has been taken into account.

Author Response

Dear reviewer,

We would like to express our gratitude for your time and valuable insights that have helped improve our research.

We have included this acknowledgment in the attached PDF document.

Best regards

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

1. The less optimal expression in line 240 is mathematically incorrect.

2. It is not clear how the population is divided into clusters. It is doubtful that the link to DBSCAN tells the reader anything.

Author Response

Dear reviewer,

We want to express our gratitude for the thorough review, which once again has allowed us to enhance important details in our research. We have addressed each of your comments in the attached PDF document. We hope to have resolved each of them.

We remain attentive to any future updates.

Best regards!

Nicolás Caselli B.

Author Response File: Author Response.pdf