Predicting Models for Plant Metabolites Based on PLSR, AdaBoost, XGBoost, and LightGBM Algorithms Using Hyperspectral Imaging of Brassica juncea

Round 1

Reviewer 1 Report

Dear authors, thank you for submitting your work to this journal. Please find some comments on your work.

 

1.- Please explain why you are referring to collect "more data", was there any previous data set collected before? Explain.

2.- Please provide references and describe the methodology used to analyze the concentration of the metabolites you are interested. Must provide the methodology including brief description of all methods involved.

3.- is it possible to add any kind of correlation to the plots of the figure 2?

4.- Seem some registration process had to be perform in the hyperspectral imagery collected, please provide information related to the image registration processing. Include RMSE for the processing and the transformation used, i.e., affinity, etc. Also provide information if any type of geometric transformation had to be performed.

5.- Please provide more critical insights on your work in the conclusions. Please provide directions for further research.

6.- Please provide more information on how you grew the plants, soil, fertilization, plant disease control, etc.

7.- Why that number of plants?

8.- Explain how irrigation was performed.

9.- why did you choose this plant to perform the study?

 

Author Response

Thanks for your comments. Please check the attached file.

Author Response File: Author Response.docx

Reviewer 2 Report

the manuscript describes an approach based on PLSR, 2 Adaboost, XGboost and LightGBM algorithms to predict plant metabolites. The line of research is current and interesting, however it requires revision to better define some points.

 

The introduction is too short. Much more information and related references on the state of the art would be useful.

In addition, a state of the art on machine learning techniques combined with hyperspectral imaging in the agro-food sector would also be needed.

Device: What device was used? measurement conditions, experimental set-up? I have not found complete information to be able to make the measurements repeatable.

 

Results:Add the mean and preprocessed spectra of the considered classes with a brief description of the absorptions.

 

Figure 3 is very interesting I suggest the authors to enlarge the images using a full page for these results. In addition, have other methods been considered to confirm the results?

 

Author Response

Thanks for your comments. Please check the attached file.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Thank you for improving the quality of your work.

 

Author Response

Thank you for the comment.

Reviewer 2 Report

the manuscript has improved significantly.

I suggest specifying in the sentence in lines 63 and 64 the possibility of identifying contaminants in food using machine learning combined with hyperspectral imaging (reference: Bonifazi, G., Capobianco, G., Gasbarrone, R., & Serranti, S. (2021). Contaminant detection in pistachio nuts by different classification methods applied to short-wave infrared hyperspectral images. Food Control, 130, 108202.)

Author Response

Thank you for the comment.

Revised the sentence and added reference as your suggestion (L63-65, 332-334).

“Machine learning techniques combined with hyperspectral imaging have been extensively used for determination of food quality [10], such as identifying contaminants in food [11]. Among them, boosting methods …”