Investigation of Passengers’ Perceived Transfer Distance in Urban Rail Transit Stations Using XGBoost and SHAP

Round 1

Reviewer 1 Report

Transfer Distance identification is vital in passenger flow management and urban rail transit station design and planning. The topic is interesting and valuable. However, there are some issues that should be enhanced.

1.      The novelty, contribution, and potential application of the paper should be highlighted.

2.      The proposed method is suggested to be compared with other state-of-the-art.

3.      It is suggested to compare the performance of the proposed method using the key factors and all factors.

Author Response

Dear Editor,
I am writing to submit the revised version of my paper titled "Investigation of Passengers' Perceived Transfer Distance in Urban Rail Transit Stations Using XGBoost and SHAP". In response to the reviewers' comments, I have made significant revisions to the manuscript, which I believe have improved its clarity and overall quality.
Along with the revised manuscript, I have also included a separate document titled "response-sustainability-2364319-0501". This document outlines the changes I made to the manuscript in response to the reviewers' comments and provides a detailed explanation of each modification.
I appreciate the thorough review provided by the reviewers and the opportunity to improve my manuscript based on their feedback. I hope that the revised version of the paper meets the requirements of the journal and that you will find it suitable for publication.
Thank you for your time and consideration.
Sincerely,
Wenjiao Xu

Author Response File: Author Response.docx

Reviewer 2 Report

 

 

The authors of this paper have presented a novel approach to study the perceived transfer distance of passengers at urban rail transit stations using XGBoost and SHAP (SHapley Additive exPlanations). The paper is well-structured and addresses a crucial aspect of transportation planning, which can help optimize passenger flow and improve the overall user experience in urban rail transit systems. The methodology is sound, and the application of machine learning techniques, specifically the XGBoost algorithm, adds to the novelty of the study.

However, there are a few areas that could be improved to strengthen the paper:

1. Literature Review: The paper would benefit from an expanded literature review, discussing more studies that have explored passenger perception of transfer distances in urban rail transit systems. This would help establish the relevance of the current study and provide context for the reader. Here are a few papers that can be incorporated into the literature review:
• Guo, Z., & Wilson, N. H. M. (2011). Assessing the cost of transfer inconvenience in public transport systems: A case study of the London Underground. Transportation Research Part A: Policy and Practice, 45(2), 91-104. doi:10.1016/j.tra.2010.12.001
• Chen, J., & Chen, X. (2019). Transfer distance, service quality, and user loyalty in urban transit systems. Journal of Transport Geography, 78, 42-54. doi:10.1016/j.jtrangeo.2019.05.012
• Alshboul, O., Shehadeh, A., Mamlook, R. E. A., Almasabha, G., Almuflih, A. S., & Alghamdi, S. Y. (2022). Prediction Liquidated Damages via Ensemble Machine Learning Model: Towards Sustainable Highway Construction Projects. Sustainability, 14(15), 9303. 

 

https://doi.org/10.3390/su14159303

 

1. Model Validation: The paper would benefit from a more thorough description of the model validation process. The authors should provide more details on the train-test split, cross-validation, and other validation techniques used to ensure the robustness and accuracy of the XGBoost model.
2. Feature Importance: While the authors mention the use of SHAP values for feature importance, it would be useful to include a detailed discussion and visualization of the feature importance results. This could help readers better understand the key factors affecting passengers' perceived transfer distances.
3. Model Interpretability: The authors should elaborate on the interpretability of the XGBoost model in the context of perceived transfer distance. Discussing the advantages and limitations of using a tree-based model, such as XGBoost, in comparison to more easily interpretable models like linear regression, would add value to the paper.
4. Implications and Applications: The paper should discuss the practical implications of the findings, such as potential improvements in station design, signage, and passenger information systems. Providing specific examples of how the study's findings could be applied in real-world urban rail transit systems would increase the paper's relevance to practitioners and policymakers.

In conclusion, the paper presents a promising approach to understanding perceived transfer distance at urban rail transit stations using the XGBoost algorithm and SHAP values. By addressing the suggested improvements, the paper can provide a more comprehensive understanding of the topic and establish its relevance to the broader transportation planning community.

Here are some suggestions:

1. In the title, consider revising it to: "Investigation of Passenger's Perceived Transfer Distance in Urban Rail Transit Stations Using XGBoost and SHAP" for better coherence.
   
2. In the abstract, some sentences seem to be a comment on the manuscript rather than a part of the abstract. Consider revising these sentences.
3. Throughout the manuscript, ensure consistency in the use of tenses. For example, when discussing the methods, use the past tense to describe the processes that have been completed. 
4. Be cautious of the use of passive voice, as it can sometimes make sentences less clear. Where possible, use active voice for better readability.
5. When presenting the results, ensure that all tables and figures are properly labeled, and the text refers to them accurately.
6. Proofread the entire manuscript to correct the major grammatical errors or inconsistencies in punctuation.

Author Response

Dear Editor,
I am writing to submit the revised version of my paper titled "Investigation of Passengers' Perceived Transfer Distance in Urban Rail Transit Stations Using XGBoost and SHAP". In response to the reviewers' comments, I have made significant revisions to the manuscript, which I believe have improved its clarity and overall quality.
Along with the revised manuscript, I have also included a separate document titled "response-sustainability-2364319-0501". This document outlines the changes I made to the manuscript in response to the reviewers' comments and provides a detailed explanation of each modification.
I appreciate the thorough review provided by the reviewers and the opportunity to improve my manuscript based on their feedback. I hope that the revised version of the paper meets the requirements of the journal and that you will find it suitable for publication.
Thank you for your time and consideration.
Sincerely,
Wenjiao Xu

Author Response File: Author Response.docx

Reviewer 3 Report

 

The data and paper is incomplete- they should describe how the survey is conducted for transferability. The paper does not provide information on the data collection methods or sample characteristics, which could affect the validity and reliability of the results. Secondly, how are the actual distances measured and how do they define whether a distance was right or wrong given that the survey gives a scalar number for distance rather than a distance band for example, 200-250m rather than 250m? A better description of the data is required for completeness of the paper.

Perceived Travel time might be an even important dependent variable for these type of studies compared to just distance.

Language is for the most part very straight forward and easy to read.

Author Response

Dear Editor,
I am writing to submit the revised version of my paper titled "Investigation of Passengers' Perceived Transfer Distance in Urban Rail Transit Stations Using XGBoost and SHAP". In response to the reviewers' comments, I have made significant revisions to the manuscript, which I believe have improved its clarity and overall quality.
Along with the revised manuscript, I have also included a separate document titled "response-sustainability-2364319-0501". This document outlines the changes I made to the manuscript in response to the reviewers' comments and provides a detailed explanation of each modification.
I appreciate the thorough review provided by the reviewers and the opportunity to improve my manuscript based on their feedback. I hope that the revised version of the paper meets the requirements of the journal and that you will find it suitable for publication.
Thank you for your time and consideration.
Sincerely,
Wenjiao Xu

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

I am delighted to share my thoughts on your revised paper, which presents a well-conducted and insightful study on the perceived transfer distance of passengers at urban rail transit stations. Your research not only fills a critical gap in the literature but also has the potential to significantly impact urban transport planning.

Here are some aspects of your paper that I found particularly noteworthy:

1. Your comprehensive literature review expertly sets the stage for your research, establishing the need for a more accurate understanding of passengers' perception of transfer distances in urban rail transit stations.
2. The utilization of the XGBoost algorithm demonstrates a cutting-edge approach to modeling and predicting perceived transfer distances. This choice of method highlights your commitment to producing reliable and robust results.
3. Your innovative application of SHAP values enables the identification of key factors that influence passengers' perceived transfer distances. This information is invaluable for urban planners and transport authorities who aim to optimize station design and enhance passenger satisfaction.
4. The clarity of your writing and presentation of results, including visualizations and tables, makes your paper accessible to readers from various backgrounds. This will undoubtedly contribute to the broader impact of your work.
5. Your thorough discussion of the results and their implications not only reinforces the value of your research but also opens up avenues for further investigation in this area.
   
   

It is vital to ensure that ALL my comments from the first round of review is taken into consideration.