Artificial Intelligence: A Promising Tool for Application in Phytopathology

Shortcomings

Response

1

Overall, the main focus of this article should be on the application of AI in detecting plant diseases, as indicated by the article title. About "Phytopathology" and artificial intelligence technology, the content is supposed to be reduced

We have extensively expanded section 5, with greater depth in the analysis of each work examined and with an increased amount of explanations about the methods used and their results. Additionally, we have included a new subsection 5.3 discussing the aforementioned methods and comparing them with each other.

2

From line 172 to line 355, a significant amount of space is devoted to introducing the fundamental knowledge of artificial intelligence. However, this information has weak relevance to the main focus of this article, which should be on plant diseases.

We agree with the reviewer that the section dedicated to the fundamentals of AI is somewhat lengthy. We believe that this information serves as a good illustration and introduction for readers not familiar with these fundamentals. Nevertheless, following the recommendations made, we have shortened its length from about 1600 words to just 1000 words, resulting in a reduction of more than 35%.

3

As a review article, this article lacks performance evaluation comparisons for different artificial intelligence methods in plant disease detection, advancements in modeling, etc.

we have included a new subsection 5.3 discussing the aforementioned methods and comparing them with each other.

4

In Section 5, "Applications of AI in Phytopathology", it is noted that "5.3 Precision Agriculture and Management" does not pertain to the topic of "Phytopathology" discussed in the article. The definition of "Phytopathology" in the text is "...scientific discipline...the study of plant diseases", as stated in line 70.

Although the Precision Agriculture and Management section is not directly related to phytopathology as indicated by the reviewer, we consider the information developed and reviewed within it to be relevant and related. Therefore, following the criteria set out by reviewer 1, this subsection has been extracted from section 5 and a new section titled 6. Applications of AI in Precision Agriculture and Management has been created, subsequently renumbering the rest of the sections according to the new layout.

5

It is suggested to add some images to represent the principles of various artificial intelligence methods visually

Two new figures have been generated to illustrate the article a bit: Figure 3. Graph diagram illustrating the operation of Siamese Convolutional Neural Networks (SNNs) for detecting plant diseases from an image dataset; and Figure 4. Graph diagram illustrating how the Random Forest system works.

Shortcomings

Response

1

Incorrect subheader format: 4.3 Relevance of AI in Various Fields, 4.3.5 Space Exploration: Autonomous Exploration and Data Analysis,

This mistakes have been corrected.

2

From my perspective, there is a disproportionate allocation of content with approximately 9.5 pages dedicated to general information about AI as opposed to 6.5 pages about Phytopathology

We agree with the reviewer that the section dedicated to the fundamentals of AI is somewhat lengthy. We believe that this information serves as a good illustration and introduction for readers not familiar with these fundamentals. Nevertheless, following the recommendations made, we have shortened its length from about 1600 words to just 1000 words, resulting in a reduction of more than 35%.

Additionally, we have extensively expanded section 5, with greater depth in the analysis of each work examined and with an increased amount of explanations about the methods used and their results. A new subsection 5.3 discussing the aforementioned methods and comparing them with each other was included.

Two new figures have been generated to illustrate the article a bit: Figure 3. Graph diagram illustrating the operation of Siamese Convolutional Neural Networks (SNNs) for detecting plant diseases from an image dataset; and Figure 4. Graph diagram illustrating how the Random Forest system works.

3

The article evidently appears to have been composed using ChatGPT

Some computer applications were used to translate some expressions and phrases from Spanish to English, including AI tools. We have conducted a thorough review of the manuscript to try to avoid some repeated words or phrases that were erroneously included as a result of these automated correction processes. Thank you for pointing this out so that the current version can be of higher quality.

This thorough review, which includes 157 references, many of them very recent, has been crafted and elaborated with great care and analysis, and based on our knowledge and scientific experience in writing original articles, reviews, and book chapters by the team of authors listed in the review.

4

It may be beneficial to consider reviewing the following papers about

We appreciate the reviewer's recommendations for potential works. We have consulted the references and found it highly interesting to include the themes of vision transformers and siamese networks. These works are extensively explained in section 5.1 between lines 394 and 469. We have even included a figure 3 about the SNNs workflow.

5

In Table 1, there is an absence of traditional machine learning approaches such as:

a. Multilayer Perceptron (MLP)

b. Stochastic Gradient Descent (SGD)

c. Random Forest

d. ElasticNet/Lasso/Ridge Regression

e. Decision Tree/XGBoost

We found the suggestion made by the reviewer very interesting and in line with that, we have updated and expanded the information in table 1 with the proposed models.

6

There is a lack of discussion on existing pretrained models that could be applied in Phytopathology.

A new subsection 5.3 discussing the aforementioned methods and comparing them with each other was included. This subsection include mention to pretrained models.

7

Is there any application of Explainable Artificial Intelligence (XAI) in Phytopathology? It would be worthwhile to engage in a discussion on this topic

The studies on Explainable Artificial Intelligence (XAI) are both intriguing and promising, yet there are currently no published studies on its application specifically within the field of phytopathology. However, XAI has seen application in various other domains, such as healthcare management, industrial asset prognostics, and electronic health record analysis, showcasing its broad potential and utility.