Spatial Autocorrelation Incorporated Machine Learning Model for Geotechnical Subsurface Modeling
Round 1
Reviewer 1 Report
It needs to clearly state in the paper the advantages of using the proposed approach of modelling (especially in the introduction). It is not clear from the introduction section why the authors propose their approach.
It is recommended to compare the results obtained with the new approach and with other approaches known from the literature (for exmple, proposed by Behrens et al.)
The authors should indicate how their approach can be used in other fields such as reservoir modelling.
There is also a problem with a proper displaying referencing in the introduction section.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 2 Report
The authors introduce a machine learning approach for geotechnical subsurface modeling incorporating spatial autocorrelation. The approach generates a new set of features called Euclidean distance fields (EDF) based on the distance between query points and observed boreholes. Principal component analysis (PCA) is applied to reduce the dimensionality of the dataset. Several machine learning algorithms (SVM, GPR, ANN, and KNN) are employed to predict geotechnical information. Results show that the EDF model performs slightly better than the model with a direct XY coordinate approach. However, I have the following comments for the authors to consider.
(1) In Figure 3, the authors should add a detailed description of the data. For example, what does the width of the histogram represent?
(2) Hyperparameter tuning is one of the research focuses, but there are too few descriptions on this part.
(3) As the authors mentioned in 2.4.3, there are two different features for location information, but in Table 3, there are not only two features for each target.
(4) The authors should add explanations to the legend in Figure 6.
(5) The authors performed experiments on EDF_80、EDF_90、EDF_NOPCA、XY_PCA, how the results will vary if use XY coordinate approach with PCA?
(6) The characters of Figure 12 are small, The authors should redesign it.
(7) There are multiple citation errors in the article, for example, ‘Error! Reference source not found’ , and in line 256, ‘3314’ does not show in references.
(8) Chapter numbers are out of order; for example, after section 3.2 is section 3.4 and there is no chapter 4 in the article. Please check whether the content of the article is missing or the serial number is wrong.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The article is recommended for publication.
Author Response
Point 1:
Reviewer 1 has a comment on the quality of English language and style which required extensive editing.
Response 1:
The editing of English language and style as suggested by reviewer 1 has been done by revising the grammar and rephrasing sentences. The changes within the manuscript related to the comment are highlighted in yellow.
Reviewer 2 Report
The paper is well written, and all my issues have been addressed. I have one small question:
Chapter numbers are out of order; for example, after chapter 3 is chapter 5.
Author Response
Point 1:
The paper is well written, and all my issues have been addressed. I have one small question:
Chapter numbers are out of order; for example, after chapter 3 is chapter 5.
Response 1:
The section number for “Conclusion” has been renumbered. The change within the manuscript related to the comment is highlighted in gray (line 554).