Identifying Traffic Congestion Patterns of Urban Road Network Based on Traffic Performance Index

Round 1

Reviewer 1 Report

The manuscript's title appears to be appropriate, but the content of the manuscript does not appear to be suitable for the journal.

The study aims to develop a clustering method to classify congestion patterns based on the traffic performance index (TPI) in Beijing. The self-organizing Maps (SOM) clustering method improved by an automatic clustering number determination algorithm is proposed to recognize congestion patterns.

The abstract section contains the necessary information. However, the results mentioned in the abstract are unclear. Please re-write the result part in the abstract.

The introduction section lacks clarity, as the need for the study is not well justified. Also, the organization of the remainder of the paper should have been provided to help understand the flow of such a long manuscript. The research gap in the literature should be made explicit, and the application of the method developed in the study should be explained.

The methodology section is too voluminous; please condense this section by eliminating less important information.

The pictorial presentation of the results looks impressive. However, the analysis results could have been explained in a bit more detail.

 

I advise the authors to have a native English speaker read and amend their paper before resubmitting it. Please see other minor comments in the attached PDF.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

  This paper presents a new method of clustering to classify congestion. The topic is interesting but some drawbacks need to be addressed before ready for publication.


1. The title does not really represent its aim. The aim is "This paper aims to develop a clustering method to classify congestion patterns based on the traffic performance index (TPI) in Beijing.", while in the title stated that the clustering is based on Improved SOM Neural Network. The author should clearly distinguish the role of TPI and SOM neural network in the method. And it should be clearly expressed in the title, abstract, introduction, results and conclusion.

2. If this paper aims to develop a new method of clustering, the Literature Review should elaborate the existing methods and present the weaknesses of them.

3. Section 3 should be part of the methodology, and the flow from development of TPI to clustering should be clearly explained.

4. The case study should also be introduced in the method.

5. A validation should be conducted on the clustering results. It can be compared with the results from the existing robust method of the same case study and analyze the accuracy.

6. The results of the study are not discussed with adequate references.

7. The role of SOM neural network is not clearly stated in the conclusion.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

I appreciate the works of the authors to revise this paper.  I believe the current form is greatly improved. However, with respect to the comparison to the existing model, the author should judge weather they have made improvements or not. I think after this sentence "The silhouette coefficient of congestion patterns in this paper is 0.94, which is higher than the reference of Li et al. [20].". What does it mean? is it better? any implication?

Author Response

Thanks for your encourage and insightful comment. The meaning of the higher silhouette coefficient is explained in the revised manuscript. (Page 15, Lines 463 to 465 in the revised manuscript) The closer the value of silhouette coefficient is to 1, the better the clustering results. The clustering accuracy of traffic congestion patterns is improved by the SOM clustering method proposed in this paper.

Author Response File: Author Response.docx