An LSTM-Based Method Considering History and Real-Time Data for Passenger Flow Prediction
Round 1
Reviewer 1 Report
This manuscript presents a novel passenger flow prediction network model based on long short-term memory networks. Using this passenger flow prediction network model in this case study is quite interesting, but it was expected to appear similar approach 'cause it is obvious. The design of this manuscript is good. There is enough results confirmation.
The "2. Related Work" section is done quite formally, there is no sequence of conclusions and motivations.
The list of references contains too many references to local authors works but there are not only this works related to manuscript' theme.
The literature review (Related Work section) can be improved by adding additional references.
There is no future work claimed, but this study required additional research caused by practical application of this work.
English proofreading is preferable.
Author Response
Dear reviewer:
Thank you for your comments of our manuscript. We have revised our manuscript based on these comments.
Point 1: The "2. Related Work" section is done quite formally, there is no sequence of conclusions and motivations.
Response 1: We have added the following content “In this part, we will mainly introduce three methods used as baselines in our experiment. We will also introduce the achievements researchers got and the shortcomings of these methods. This part also gives the reason why we increase the data dimensions through POI data and use two encoder structures in our model based on LSTM.” from line 63 to line 66 to explain the propose of “Related Work” part.
Point 2 and point 3: The list of references contains too many references to local authors works but there are not only this works related to manuscript' theme.
The literature review (Related Work section) can be improved by adding additional references.
Response 2 and 3: We have replaced some references based on the comments. The revised content and new references in “Related Work” are as follows:
“R. Gummadi et al. [2] chose time series model to forecast the passenger flow for a certain transit bus station.” in line 71.
“S. V. Kumar et al. [5] used Kalman filtering to predict the traffic flow.” in line 75.
“F. Toqué et al. [13] used LSTM neural networks to predict travel demand based on smart card data and proved the effectiveness of the forecasting approaches. S. Sunardi et al. [14] presented a useful way of observing the domestic passenger travel series using LSTM.” from line 98 to line 100.
“K. Pasini et al. and M. Farahani et al. [19,20] used encoder structures based on LSTM to predict passenger number and draw a conclusion that an encoder structure can improve the accuracy. That is the main reason that we use two encoder structures in our model.” from line 111 to line 114.
[2] R. Gummadi and S. R. Edara, "Analysis of Passenger Flow Prediction of Transit Buses Along a Route Based on Time Series," in Information and Decision Sciences: Springer, 2018, pp. 31-37.
[5] S. V. Kumar, "Traffic flow prediction using Kalman filtering technique," Procedia Engineering, vol. 187, pp. 582-587, 2017.
[13] F. Toqué, M. Khouadjia, E. Come, M. Trepanier and L. Oukhellou, "Short & long term forecasting of multimodal transport passenger flows with machine learning methods," 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC), Yokohama, 2017, pp. 560-566, doi: 10.1109/ITSC.2017.8317939.
[14] S. Sunardi, D. Dwiyanto, M. Sinambela, J. Jamaluddin, and D. R. Manalu, "Prediction of Domestic Passengers at Kualanamu International Airport Using Long Short Term Memory Network," MEANS, vol. 4, no. 2, pp. 165-168, 2019.
[19] K. Pasini, M. Khouadjia, A. Same, F. Ganansia, and L. Oukhellou, "LSTM encoder-predictor for short-term train load forecasting," 2019.
[20] M. Farahani, M. Farahani, M. Manthouri, and O. Kaynak, "Short-Term Traffic Flow Prediction Using Variational LSTM Networks," arXiv, 2020.
Point 4 and point 5: There is no future work claimed, but this study required additional research caused by practical application of this work.
English proofreading is preferable.
Response 4 and 5: We added the following context “In the future, we will try to find a more proper sequence length for both history data encoding part and real-time encoding part to make our research more completely. We hope this research can help the bus operation and improve the bus system stability.” from line 372 to line 375 as our future work. We also revised some grammar errors in our paper.
Yours
Qi Ouyang
Author Response File: Author Response.pdf
Reviewer 2 Report
The authors of the paper present an LSTM-based method considering history and real-time data for passenger flow prediction (HRPFP model to predict passenger numbers for each station at a certain line using the history data and the real-time collected passenger number). The subject of the paper is interesting, worth investigating, and is of great importance in the operational management of public transport.
The authors present an overview of recent years' research using neural networks and deep learning to passenger flow prediction.
The research methods and algorithms used seem to be properly selected for the research.
The results are presented clearly and comprehensibly. The conclusions were supported by the research results
Author Response
Dear reviewer:
Thank you for reviewing our manuscript.
Yours
Qi Ouyang
Author Response File: Author Response.pdf