The Journal of The Korea Institute of Intelligent Transport Systems (한국ITS학회 논문지)

The Korea Institute of Intelligent Transport Systems (한국ITS학회)
권호 표지
DOI QR코드

DOI QR Code

Time Series Modeling Pipeline for Urban Behavioral Demand Prediction under Uncertainty

COVID-19 사례를 통한 도시 내 비정상적 수요 예측을 위한 시계열 모형 파이프라인 개발 연구

  • Minsoo Jin (Regional Cooperation & Promotion Division, Korea Institute of Civil Eng. and building Technology) ;
  • Dongwoo Lee (Dept. of Urban Planning, Univ. of Incheon National) ;
  • Youngrok Kim (Dept. of Highway & Transportation Research, Korea Institute of Civil Eng. and building Technology) ;
  • Hyunsoo Lee (Dept. of Urban Planning, Univ. of Incheon National)
  • 진민수 (한국건설기술연구원 지역협력진흥실) ;
  • 이동우 (인천대학교 도시행정학과) ;
  • 김영록 (한국건설기술연구원 도로교통연구본부) ;
  • 이현수 (인천대학교 도시행정학과)
  • Received : 2023.03.15
  • Accepted : 2023.04.26
  • Published : 2023.04.30
Download PDF
⟨ Previous Next ⟩

Abstract

As cities are becoming densely populated, previously unexpected events such as crimes, accidents, and infectious diseases are bound to affect user demands. With a time-series prediction of demand using information with uncertainty, it is impossible to derive reliable results. In particular, the COVID-19 outbreak in early 2020 caused changes in abnormal travel patterns and made it difficult to predict demand for time series. A methodology that accurately predicts demand by detecting and reflecting these changes is, therefore, required. The current study suggests a time series modeling pipeline that automatically detects and predicts abnormal events caused by COVID-19. We expect its wide application in various situations where there is a change in demand due to irregular and abnormal events.

도시에 많은 사람들이 밀집하여 살아가면서 기존에 예측하지 못했던 범죄, 사고, 감염병 등의 비정상 이벤트가 발생은 도시 내 이용자 수요에 영향을 미치게 된다. 이러한 불확실성(uncertainty)이 내포된 정보를 기반으로 도시 내 이용자 수요에 대한 시계열적 예측을 수행한다면 신뢰성 있는 결과 도출이 불가능하다. 특히, 2020년 초 발발한 COVID-19는 비정상적인 이동통행패턴의 변화를 불러 일으키며 시계열 수요예측을 어렵게 만들었기에 이러한 변화를 검지하고 이를 반영하여 정확한 수요를 예측 수행할 수 있는 방법론의 필요성이 대두되고 있다. 이에 본 연구는 COVID-19로 인한 비정상적 이벤트를 자동으로 검지하고 예측하는 모형 파이프라인을 구축하였다. 이는 도시 내 다양한 분야에서의 불규칙적이고 비정상적인 이벤트로 인한 수요변화가 일어나는 상황에 폭넓게 활용될 수 있을 것으로 생각된다.

Keywords

Acknowledgement

본 연구는 국토교통부 "AI·데이터 기반 스마트시티 통합플랫폼 모델 개발 및 실증연구"의 연구비지원(과제번호: 22AIIP-C163095-02)에 의해 수행되었습니다.

References

  1. Bajari, P., Nekipelov, D., P.Ryan, S. and Yang, M.(2015), "Machine Learning Methods for Demand Estimation", American Economic Review, vol. 105, no. 5, pp.481-485. https://doi.org/10.1257/aer.p20151021
  2. Cha, D. H., Lee, K. H., Son, J. H., Sun, D. W., Lim, H. T., Lee, J. K., Choi, T. J. and Goo, Y. M.(2021), "Prediction of Breeding Animals by Swine Breeding Scale using ARIMA Model in Time Series Analysis", Journal of Animal Breeding and Genomics, vol. 5, no. 2, pp.31-50.
  3. Han, D. C., Lee, D. W. and Jung, D. Y.(2021), "Estimation of Shared Bicycle Demand Using the SARIMAX Model: Focusing on the COVID-19 Impact of Seoul", Journal of the Korea Institute of Intelligent Transportation Systems, vol. 20, no. 6, pp.1-13. https://doi.org/10.12815/kits.2021.20.6.1
  4. Han, D., Lee, D. W. and Jung, D.(2021), "A Study on the Traffic Volume Correction and Prediction Using SARIMA Algorithm", The Journal of the Korea Institute of Intelligent Transport Systems, vol. 20, no. 6, pp.1-13. doi: 10.12815/kits.2021.20.6.1
  5. Hong, J. Y., Han, E. R., Choi, C. H., Lee, M. S. and Park, D. J.(2021), "Estimation of Shared Bicycle Demand Using the SARIMAX Model: Focusing on the COVID-19 Impact of Seoul", Journal of The Korea Institute of Intelligent Transportation Systems, vol. 20, no. 1, pp.10-21. https://doi.org/10.12815/kits.2021.20.1.10
  6. Kim, D. K. and Kim, H. Y.(2019), "Forecasting Urban Railway Demand of Daegy Based on Time Series Analysis", Journal of Decision Making Research, vol. 27, no. 1, pp.1-11.
  7. Kim, D. K.(2020), "Urban Railway Demand Forecast using Time Series Data", Journal of the Korean Data Analysis Society, vol. 22, no. 2, pp.753-765. https://doi.org/10.37727/jkdas.2020.22.2.753
  8. Kim, J. H. and Kim, J. Y.(2022), "Prediction of Covid-19 confirmed number of cases using SARIMA model", Journal of Korea Institute of Information and Communication Engineering, vol. 26, no. 1, pp.58-63.
  9. Kim, J. S., Gang, J. H., Kim, S. H. and Youn, J. S.(2020), "A Univariate Time Series Forecasting with Prophet", Journal of Korea Institute of Information and Communication Engineering, vol. 24, no. 1, pp.329-331.
  10. Korea Transportation Safety Authority and Ministry of Land and Transport(2021), Investigation of Public Transportation System.
  11. Lim, S. H.(2020), "An Analysis of Change in Traffic Demand with Coronavirus Disease 2019", Journal of the Korea Institute of Intelligent Transportation Systems, vol. 19, no. 5, pp.106-118. https://doi.org/10.12815/kits.2020.19.5.106
  12. Noh, Y. S. and Do, M. S.(2015), "Short-term Railway Passenger Demand Forecasting by SARIMA Model", Journal of the Korea Institute of Intelligent Transportation Systems, vol. 14, no. 4, pp.18-26. https://doi.org/10.12815/kits.2015.14.4.018
  13. Oh, S. W., Lim, N. H., Lee, S. H. and Kim, M. S.(2020), "Long-term Price Prediction and Trend Analysis of Garlic Using Prophet Model", Journal of the Korean Data Analysis Society, vol. 22, no. 6, pp.2325-2336. https://doi.org/10.37727/jkdas.2020.22.6.2325
  14. Seo, C. B. and Sung, M. J.(2022), "The Impact of COVID-19 on Transportation Demand-OO Private Expressway", Proceedings of the KOR-KST Conference, pp.269-274.
  15. Shin, D. R., Chae, G. Y. and Park, M. J.(2022), "Prediction of COVID-19 Confirmed Cases by Using Big Data and Time Series Analysis", Journal of Applied Reliability, vol. 22, no. 4, pp.352-362. https://doi.org/10.33162/JAR.2022.12.22.4.352
  16. Winters, P. R.(1960), "Forecasting sales by exponentially weighted moving averages", Management Science, vol. 6, no. 3, pp.324-342. https://doi.org/10.1287/mnsc.6.3.324