Earth Observation Satellite Imagery Information Based Decision Support Using Machine Learning

Round 1

Reviewer 1 Report

 

Review of “Earth Observation Satellite Imagery Information based Decision Support using Machine Learning”

Journal: Remote Sensing

Manuscript Number: remotesensing-1794870

Author: Bruno Ferreira1, Rui G. Silva and Muriel Iten

 

            The review article is very interesting and reviewed EO and ML that handle EO data and addresses a variety of challenges across different fields. Overall, the review entitled “Earth Observation Satellite Imagery Information based Decision Support using Machine Learning” is well written, however, before publication, I would like to recommend minor changes to be included.

Minor concerns:

1.     In Figure1, Chapter 3, chapter 4, and chapter 5 should be replaced by section 3, section 4 and section 5 respectively.

2.     Line 175: “Error! Reference source not found” should be replaced by Figure 3, I guess. Linking error. Please correct it.

3.     Figure 4 is pasted twice at line 194 and line 200. Please check.

4.     It would be best if you can include all the basic machine learning algorithms with equations for better understanding of readers.

After the recommended minor changes, the article is well suited to be published in Remote Sensing Journal.

 

Comments for author File: Comments.pdf

Author Response

The provided comments are much appreciated and will significantly improve the manuscript. Amendments have been done accordingly with the comments provided by the reviewer. All changes have been highlighted in yellow throughout the manuscript.

 

Reviewer 1:

The review article is very interesting and reviewed EO and ML that handle EO data and addresses a variety of challenges across different fields. Overall, the review entitled “Earth Observation Satellite Imagery Information based Decision Support using Machine Learning” is well written, however, before publication, I would like to recommend minor changes to be included.

Minor concerns:

1. In Figure1, Chapter 3, chapter 4, and chapter 5 should be replaced by section 3, section 4 and section 5 respectively.
   1. For a better clarification, the authors have decided to remove the references to chapters.

 

2. Line 175: “Error! Reference source not found” should be replaced by Figure 3, I guess. Linking error. Please correct it.
   1. This has been amended, the reference has been replaced by "Figure 3", as suggested.

 

1. Figure 4 is pasted twice at line 194 and line 200. Please check.
   1. Figure 4 has been removed from line 194.

 

4. It would be best if you can include all the basic machine learning algorithms with equations for better understanding of readers.
   1. The vast majority, if not all, of machine learning algorithms are mostly represented with pseudo-codes, similar to a step-by-step procedure. The manuscript presents more than 50 different machine learning algorithms, making it challenging (i.e., size of the manuscript) to add the pseudo-code for every algorithm. Hence, in addition to the references of the studies in which such algorithms have been applied and presented (Section - Machine Learning Algorithms applied to Earth Observation Data”), an other example (i.e., new Figure 4) has been in added in Section 4.2. - Machine Learning Algorithms categorisation”. This corresponds to a pseudo-code along with the following text:

"Therefore, a MLA should be represented by a pseudo-code where each step of the MLA implementation is well defined. As an example, Figure 4 presents the pseudo-code of a Decision Tree (i.e., ID3).”

Reviewer 2 Report

The manuscript is comprehensive, timely and well-written and can be considered for publication. I have some suggestion for improving the presentation of the revised version as following:

2. Introduction is very short for a review paper. please add more context about and background of the study for ML classification and EO and explain the Gap for review statement.

2. Please use the colored version for Figure 1. There are chapter 3 to 5 at the bottom of figure what do you mean? please remove them. 

3. I think it would be better to name the section 2. Literature review rather than materials and methods.

4. Figure 2 needs to be redraw and convert to colored version. Please increase the resolution too. 

5. Figure 3 needs to be redraw and convert to colored version. Please increase the resolution too.

6. Figure 4 needs to be redraw and convert to colored version. Please increase the resolution too. Additionally, you must assigned A and B for each sub-image.

7. Discussion must be separated from conclusions. It is very important that the authors provide an appropriate discussion for the review in this section. 

 

Author Response

The provided comments are much appreciated and will significantly improve the manuscript. Amendments have been done accordingly with the comments provided by the reviewer. All changes have been highlighted in yellow throughout the manuscript.

 

Reviewer 2:

The manuscript is comprehensive, timely and well-written and can be considered for publication. I have some suggestion for improving the presentation of the revised version as following:

1. Introduction is very short for a review paper. please add more context about and background of the study for ML classification and EO and explain the Gap for review statement.
   1. The authors have been added new references (i.e., [1-5]) and more context regarding:
      • Overall Background (lines 28 to 33)
      • EO (lines 40 to 44)
      • ML (lines 49 to 50 and 53)
      • Knowledge Gap and Research Question (lines 59 to 61 and 62-64).
2. Please use the colored version for Figure 1. There are chapter 3 to 5 at the bottom of figure what do you mean? please remove them. 
   1. Figure 1 has been replaced by a colored version. For a better clarification, chapters 3 to 5 (at the bottom of figure) have been removed.
3. I think it would be better to name the section 2. Literature review rather than materials and methods.
   1.  Section 2 (Methods and Materials) describes how the research (literature review) was conducted, while section 3 and 4 are the literature review (concerning EO and ML, respectively). Thus, and considering the pertinence of the comments, the authors have merged Section 2, 3 and 4, into ony one (the new Section 2 – Literature Review). The old section 3 is now section 2.1 and the old section 4 is now section 2.2. Additionally, all remaining sections were amended accordingly.
4. Figure 2 needs to be redraw and convert to colored version. Please increase the resolution too. 
   1. Figure 2 has been amended accordingly.
5. Figure 3 needs to be redraw and convert to colored version. Please increase the resolution too.
   1. Figure 3 has been amended accordingly.
6. Figure 4 needs to be redraw and convert to colored version. Please increase the resolution too. Additionally, you must assigned A and B for each sub-image.
   1. Figure 4 has been amended accordingly. This reference was also incorrectly mentioned twice in line 194, and repeated in line 200, and thus removed.. This eliminates the need of assign A and B to each sub-image.
7. Discussion must be separated from conclusions. It is very important that the authors provide an appropriate discussion for the review in this section. 
   1. A section 4 has been added and named Discussion and a section 5 named Conclusions. Also, their reference has been added in Section 1, concerning the structure of the manuscript). Additionally, more context and information was added to section 6 (lines 297 to 300 and 313-319).

2.  

Reviewer 3 Report

The authors have made a significant effort to make a comprehensive review related to the application of machine learning algorithms in Earth observation data modeling and information extraction. Although the field of ML applications addressed by the authors is quite broad, the authors managed to highlight new and significant ML trends and compare them through a series of carefully selected and representative references. The part of the paper related to the final discussion and conclusions is written in a consistent manner.

 

Author Response

The provided comments are much appreciated and will significantly improve the manuscript. Amendments have been done accordingly with the comments provided by the reviewers. All changes have been highlighted in yellow throughout the manuscript.

 

Reviewer 3:

The authors have made a significant effort to make a comprehensive review related to the application of machine learning algorithms in Earth observation data modeling and information extraction. Although the field of ML applications addressed by the authors is quite broad, the authors managed to highlight new and significant ML trends and compare them through a series of carefully selected and representative references. The part of the paper related to the final discussion and conclusions is written in a consistent manner.

 

Round 2

Reviewer 2 Report

Accept in present form