Public Transportation Operational Health Assessment Based on Multi-Source Data

Round 1

Reviewer 1 Report

The article proposes a machine learning approach to analyze public transport data. While the approach is well presented and almost interesting, I lack generalization discussion to demonstrate how results can be easily generalized to other datasets. 

Minors: 

• avoid splitting tables
• proofread the article. For instance, line 439 refers to a table, but the table number is missing.
• I appreciate that Figure 13 is also for color blind people, but the pattern attached to the blue and green colors should be different to achieve this goal properly.

Author Response

Thank you for your careful and detailed review work. We have carefully understood and responded to each of your comments and have made the corresponding corrections in the paper. Please check our response report and the revised paper. Thank you again for your comments, which have been very helpful to us.

Author Response File: Author Response.docx

Reviewer 2 Report

The reviewed article requires some corrections before being published in the journal. 1. Reference to world knowledge is insufficient. The literature review (introduction) is at a very low level. The authors should proceed with the elaboration of this issue. 2. The layout and structure of the text is incorrect. There is double numbering of the drawings in many places, no introduction to the drawings, no reference to the formulas used, etc. 3. Lack of a full description of the axis in the drawings, 4. Conclusions of the work performed are general Not supported by the results. It is not listed. 

Author Response

Thank you for your careful and detailed review work. We have carefully understood and responded to each of your comments and have made the corresponding corrections in the paper. Please check our response report and the revised paper. Thank you again for your comments, which have been very helpful to us.

Author Response File: Author Response.docx

Reviewer 3 Report

A brief summary:  

This manuscript assesses the public transportation operational health based on multiple source data by supervised machine learning methods in a simplified model, in which only a one-way bus route is considered. The authors defined the major factors affecting the bus operation and rank them by a random forest regression model. I think the authors did a good job in setting up the experiments appropriately and evaluating the machine learning methods. The results in this manuscript should be shared in our applied science community to enhance our understanding of supervised machine learning algorithms. The following are my detailed comments and I hope that those would be helpful to improve the quality of this manuscript.  

 

Broad comments:  

1. The terminology and the order numbers of figures are in chaos. For instance, road delay, roadway delay, section delay, and segment delay seem to be the same thing in this manuscript but they are used randomly. Since the concepts are defined in Section 2, they should be used consistently in the context to avoid confusion. The order numbers of the figures are repeated in this manuscript and need correcting. 
2. I recommend the authors rewrite the abstract. The purpose of an abstract is to summarize the key points of your research concisely. It should highlight the key content areas, your research purpose, the importance of your research, your methods, and last but not least, your main outcomes. The current abstract is about the structure of your paper, which reads like part of an introduction instead of an abstract. 
3. English needs some attention. I recommend the authors check the entire manuscript for English. I list a few places where the English is in particular need of improvement.  

 

Specific comments: 

L36: have -> has 

 

L60: What does DEA stand for? 

 

L87: What does SVM stand for? 

 

L137-139: “If you cannot..., you cannot...” -> If the delay of bus vehicles is not reduced through signalized intersections, bus operational health cannot be improved effectively.  

 

L189-190: Are section delay and segment delay the same thing? Can you make the definition consistent if so? 

 

L194: It seems that “route delay” appears here for the first time in Figure 4. Can you define it? 

 

L213-216: repeated 

 

L227: Table numbers are missing. 

 

L232-235: It is not necessary for these two sentences to be two independent paragraphs.  

 

L261: Why do you restart counting the number of the figures here? You already have Figure 1 before.  

 

L292: learing -> learning 

 

L295: by -> with 

 

L334: The table number is missing.  

 

L356: Figure 10? 

 

L366: proposed -> was proposed  

 

L410-412: What is the figure between the two lines? 

 

L421: Table? 

 

L427: Table? 

 

L439: Table?  

 

L448: You are using different names in the figure: road delay, cross delay, and station delay. What are they? Again, you need to make your terminology consistent throughout the manuscript.  

 

Author Response

Thank you for your careful and detailed review work. We have carefully understood and responded to each of your comments and have made the corresponding corrections in the paper. Please check our response report and the revised paper. Thank you again for your comments, which have been very helpful to us.