Evolutionary Statistical System Based on Novelty Search: A Parallel Metaheuristic for Uncertainty Reduction Applied to Wildfire Spread Prediction

Round 1

Reviewer 1 Report

The article is interesting, the authors present the use of NS for a Data Driven problem in combination with other mechanisms and a parallel model.

In general terms the paper is good, however, I have the following comments:

- What is the novelty of this paper in comparison with the following papers?

https://ieeexplore.ieee.org/abstract/document/9835372

https://www.sciencedirect.com/science/article/abs/pii/S1877750314001628?via%3Dihub

- Regarding the previous comment, authors need to clarify the differences and novelty.

- The methodology is clear, but could you please provide an extended explaination of the parallel process.

- Did you compare your method with other parallel evolutionary approaches?

- Could you please explain how the statistical methodology helps the NS? 

- Please discuss and analyze the complexity of your proposal.

Some minor comments:

- Figures should be placed before they were mentioned in text, check for examople figure 1 and 2.

- The axes of some figures are in lowercase, in my opinion the first letter of the axis should be in uppercase.

Author Response

The response is in the attached file.

Author Response File: Author Response.docx

Reviewer 2 Report

The paper reports the development of an evolutionary-based solution to solve the problem of wildfire spread prediction. The main aim of the reported study was to use the Novelty Search paradigm in order to avoid local optima and to maximize exploration.

This paper deals with an interesting subject and could be of interest for many readers. However, further explanations should be added and some corrections need to be done in order to accept this work for publication. The main concerns are listed below

-       the novelty of the paper is rather shallow; anyway, the authors should emphasize their contribution to the body of knowledge

-       Figure 1 is not properly explained

-       Is dist defined by (2) a distance measure, as it is stated in line 434?

-       Explain equation (3) - the cardinal function is missing from (3)

-       The functions used to describe Algorithm 1 are not properly explained

-       Explain how you set the value of the parameter fThreshould

-       Figures 5-9 are very informative, but the reader should figure out what it’s all about; the multiple ANOVA test is not even mentioned in the text

Author Response

The response is in the attached file.

Author Response File: Author Response.docx

Reviewer 3 Report

In this paper, a new parallel metaheuristic approach for uncertainty reduction is applied to the problem of wildfire propagation prediction. This work is timely and well-written and organized. Comments to the authors:

1) Explain the block CS-Master in detail in the revised paper.

2) What is the procedure followed in the calibration stage. Explain in detail.

3) Explain the suitable reference for equation (1).

4) How the surface slope is determined in this paper.

5) Explain the reason for selecting GA in this paper. Comment on the computational time required in this work.

6) There are some grammatical errors and typos that should be corrected before publication.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

 

The authors have applied novelty search based genetic algorithm in the optimization stage in a wildfire prediction system. The proposed approach was used in fireSim simulator. The paper is well-structured. The research problem is formulated based on related research review. The authors presented the results of experiments showing the advantages of the proposed approach over other methods.

Please improve the paper addressing the following issues:

1. Can the proposed parallel metaheuristic approach for uncertainty reduction be applied to other problems?

2. What are the limitations of its applicability?

3. What are the weak points of the proposed approach?

3. The English language should be improved when it comes to grammar and style.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

My previous comments were not resolved correctly.

The answers are inappropriate.

The authors need to clarify the differences of their own previous work with this proposal. Also, an analysis of the complexity of the method is needed to understand if it is computationally expensive.  

Reviewer 2 Report

 The manuscript has been sufficiently improved and could be published in its current version.

Regarding the authors' question regarding the figures: the authors should first state what is depicted in each figure and then comment accordingly.