Machine Learning Applications in Surface Transportation Systems: A Literature Review
Round 1
Reviewer 1 Report
This paper introduces the application of machine learning (ML) in surface transportation system (STS). Through the collection of papers, it is demonstrated that ML technology can improve social STS by considering external factors and spatial factors, taking problem definition, data accessibility and available knowledge as the three algorithm selection criteria. However, it mainly focuses on prediction problems, and lacks publicly available STS data, professional spatiotemporal ML modeling and STS problem definition expression.
My comments are provided in the following:
1.What is objective and significance of this paper? It is not clearly presented in the “introduction” section;
2. The author needs a normative, clear and reasonable summary in the “introduction” section to the difference between deep learning and traditional machine learning in STS application;
3. The author should fully analyze and discuss the advantages and disadvantages of various deep learning and machine learning;
4. The recent development of this field is not listed in the method part, and bibliometric analysis is missing;
5. The author should provide more details as to how the three central standards choose the appropriate ML algorithm for specific STS applications;
6. LSTM effectively solves the problem of gradient explosion or disappearance of cyclic neural network, but how to set the amount of historical information in the consideration of more internal and external factors to predict STS goals?
7. In the summary section, the authors point out that at present, there are only limited ML algorithms for a narrow set of STS applications. Why did different applications employ different algorithms? What are the advantages and disadvantages of ML algorithms employed in the literatures?
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Reviewer 2 Report
The authors presented a review paper on Machine Learning Application in Surface Transportation Systems.
The paper I generally well prepared and can be accepted after minor revision:
Introduction is to be extended.
The novelty of the paper is to be clearly stated
A bibliometric study is to be performed; statistics of the published paper related to the subject based on Scopus and/or web of science database.
Line 301; there is a problem with the reference.
Some figures have very low resolution.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The problems have been revised.