All Issue

2024 Vol.42, Issue 3 Preview Page

Article

Real-time Traffic Incident Clearance Time Prediction by Latent Class with Random Effect: Approach for Applying Imbalanced Data

비관측 이질성이 있는 잠재 계층별 실시간 교통사고 처리시간 예측 : 불균형 데이터 적용을 위한 접근

Songha LEE1
,
Juneyoung PARK2*
이 송하1
,
박 준영2*
1The Ph.D. Course, Smart City Engineering, Hanyang University, Ansan 15588, Korea
2Associate Professor, Department of Transportation & Logistics Engineering/Smart City Engineering, Hanyang University, Ansan 15588, Korea
1한양대학교 스마트시티공학과 석박사통합과정
2한양대학교 교통·물류공학과/스마트시티공학과 부교수
*Corresponding Author
30 June 2024. pp. 370-384
PDF XML Citation
Abstract

Minimizing the negative impacts of traffic crashes after they occur is crucial for transportation agencies. Especially, accurate traffic incident clearance time prediction, among post-traffic crash management strategies, can reduce congestion by evaluating the crash impact area and providing route guidance strategies on the predicted time. Currently, several approaches have been developed various methodologies for estimating traffic incident clearance time and explaining the relationship between the time and exploratory factors. However, most of the studies were conducted overseas, making it challenging to apply them to the Korean highway road segments, which have different geometric and traffic conditions, unlike the proposed methodology. Moreover, the previous studies utilized historical data rather than dynamic data and exploratory studies about the relationship between real-time characteristics and clearance time are limited. As such, for predicting accuracy with dynamic data, novel approach is proposed for combining latent class analysis with unobserved heterogeneity and boosting algorithms, addressing with various environmental factors. Furthermore, by implementing sampling techniques and explainable artificial intelligence to resolve the imbalance clearance time data with a long tail distribution, model performance was improved and reliable results were revealed. As a result, the latent class-based prediction model that synthesized the time data with unobserved heterogeneity showed the best performance, and spatio-temporal differences in real-time traffic characteristics, trucks, and lighting facilities were highly correlated with the time variables. In the future, the results are expected to more accurately predict the range of clearance times, which is insufficient compared to general traffic, and support the establishment of systematic traffic crash management strategies based on real-time traffic characteristics and history data.

Keywords
ensemble model
imbalanced data
latent class analysis with random effect
real-time data
traffic incident clearance time

신속한 교통사고 처리 대응을 통해 부정적인 영향을 최소화하는 것은 교통 기관에 매우 중요하다. 특히, 교통사고 대응 전략 중 정확한 사고 처리시간 예측은 사고 영향 범위를 평가하고 예측 결과에 따른 경로 안내 전략을 제공함으로써 사고로 인한 혼잡을 줄일 수 있다. 이를 위해 다양한 연구들이 수행되지만 대부분 국외 연구로, 제안하는 방법론과 다른 기하학적, 교통 조건을 가지고 있는 국내 교통환경에 적용하기에는 어려움이 있고, 동적 데이터보다는 과거에 집계한 이력 데이터를 활용하여 실시간으로 변화하는 교통 특성과 사고 처리시간에 관련된 탐색적인 연구는 제한적이다. 따라서 본연구는 다양한 환경요인을 고려하여 임의 효과를 가진 잠재 계층 분석과 부스팅 계열 알고리즘을 결합하여 실시간 사고 처리시간 예측 모형을 개발하였다. 또한, 긴 꼬리 분포를 가진 사고 처리시간 데이터의 불균형을 해결하는 샘플링 기법과 설명가능한 인공지능을 적용하여 모형 성능 개선 및 신뢰성 있는 연구 결과를 도출하였다. 연구 결과, 관찰되지 않은 이질성이 포함된 사고 처리시간 데이터를 합성한 잠재 클래스 기반 예측 모델이 가장 성능이 우수하였고 실시간 교통류의 시공간적 차이, 화물차, 미작동한 조명시설이 사고 처리시간과 높은 상관관계가 있음을 보여주었다. 이 연구 결과는 일반 교통류에 비해 부족한 사고 교통류의 다양한 사고 처리시간 크기를 더욱 정확하게 예측하고 실시간성 교통 특성 및 이력 자료를 기반으로 체계적인 교통사고 대응 전략 수립을 지원할 수 있을 것으로 기대한다.

키워드
앙상블 모형
불균형적인 데이터
임의 효과 기반 잠재 계층분석
실시간성 데이터
교통사고 처리시간
References
1

Beath K. J. (2017), RandomLCA: An R Package for Latent Class with Random Effects Analysis, Journal of Statistical Software, 81, 1-25.

10.18637/jss.v081.i13
2

Branco P., Torgo L., Ribeiro R. P. (2017), SMOGN: A Pre-processing Approach for Imbalanced Regression, In First International Workshop on Learning with Imbalanced Domains, Theory and Applications, 36-50, PMLR.

3

Branco R., Ribeiro P., Torgo L. (2016a), UBL: An R Package for Utility-based Learning, ArXiv Preprint ArXiv:1604.08079.

4

Chen T., Guestrin C. (2016), Xgboost: A Scalable Tree Boosting System, In Proceedings of the 22nd Acm sigkdd International Conference on Knowledge Discovery and Data Mining, 785-794.

10.1145/2939672.2939785
5

Chung Y., Kim S. (2020), Identification of Critical Factors for Crash Duration with Consideration of Unobserved Heterogeneity, J. Korean Soc. Transp., 38(6), Korean Society of Transportation, 536-546.

10.7470/jkst.2020.38.6.536
6

De Ona J., López G., Mujalli R., Calvo F. J. (2013), Analysis of Traffic Accidents on Rural Highways using Latent Class Clustering and Bayesian Networks, Accident Analysis & Prevention, 51, 1-10.

10.1016/j.aap.2012.10.01623182777
7

Depaire B., Wets G., Vanhoof K. (2008), Traffic Accident Segmentation by Means of Latent Class Clustering, Accident Analysis & Prevention, 40(4), 1257-1266.

10.1016/j.aap.2008.01.00718606254
8

Haule H. J., Sando T., Lentz R., Chuan C. H., Alluri P. (2019), Evaluating the Impact and Clearance Duration of Freeway Incidents, International Journal of Transportation Science and Technology, 8(1), 13-24.

10.1016/j.ijtst.2018.06.005
9

Hojati A. T., Ferreira L., Washington S., Charles P. (2013), Hazard Based Models for Freeway Traffic Incident Duration, Accident Analysis & Prevention, 52, 171-181.

10.1016/j.aap.2012.12.03723333698
10

Hou L., Lao Y., Wang Y., Zhang Z., Zhang Y., Li Z. (2014), Time-varying Effects of Influential Factors on Incident Clearance Time using a Non-proportional Hazard-based Model, Transportation Research Part A: Policy and Practice, 63, 12-24.

10.1016/j.tra.2014.02.014
11

Islam N., Adanu E. K., Hainen A. M., Burdette S., Smith R., Jones S. (2021), A Comparative Analysis of Freeway Crash Incident Clearance Time using Random Parameter and Latent Class Hazard-based Duration Model, Accident Analysis & Prevention, 160, 106303.

10.1016/j.aap.2021.10630334303495
12

Jeihani M., James P., Saka A. A., Ardeshiri A. (2015), Traffic Recovery Time Estimation under Different Flow Regimes in Traffic Simulation, Journal of Traffic and Transportation Engineering (English Edition), 2(5), 291-300.

10.1016/j.jtte.2015.08.001
13

Khattak A., Wang X., Zhang H. (2012), Incident Management Integration Tool: Dynamically Predicting Incident Durations, Secondary Incident Occurrence and Incident Delays, IET Intelligent Transport Systems, 6(2), 204-214.

10.1049/iet-its.2011.0013
14

Kuang, L., Yan, H., Zhu, Y., Tu, S., Fan, X. (2019), Predicting Duration of Traffic Accidents Based on Cost-sensitive Bayesian Network and Weighted K-nearest Neighbor, Journal of Intelligent Transportation Systems, 23(2), 161-174.

10.1080/15472450.2018.1536978
15

Lee S. B., Han D. H., Lee Y. I. (2015), Development of Freeway Traffic Uncident Clearance Time Prediction Model by Accident Level, J. Korean Soc. Transp., 33(5), Korean Society of Transportation, 497-507.

10.7470/jkst.2015.33.5.497
16

Li R., Pereira F. C., Ben-Akiva M. E. (2018), Overview of Traffic Incident Duration Analysis and Prediction, European Transport Research Review, 10(2), 1-13.

10.1186/s12544-018-0300-1
17

Li, R. (2015), Traffic Incident Duration Analysis and Prediction Models Based on the Survival Analysis Approach, IET Intelligent Transport Systems, 9(4), 351-358.

10.1049/iet-its.2014.0036
18

Lundberg S. M., Lee S.-I. (2017), A Unified Approach to Interpreting Model Predictions, Advances in Neural Information Processing Systems, 30, 4768-4777, Red Hook, NY, Curran Associates.

19

Ma X., Ding C., Luan S., Wang Y., Wang Y. (2017), Prioritizing Influential Factors for Freeway Incident Clearance Time Prediction using the Gradient Boosting Decision Trees Method, IEEE Transactions on Intelligent Transportation Systems, 18(9), 2303-2310.

10.1109/TITS.2016.2635719
20

Mannering F. L., Shankar V., Bhat C. R. (2016), Unobserved Heterogeneity and the Statistical Analysis of Highway Accident Data, Analytic Methods in Accident Research, 11, 1-16.

10.1016/j.amar.2016.04.001
21

Nam D., Mannering F. (2000), An Exploratory Hazard-based Analysis of Highway Incident Duration, Transportation Research Part A: Policy and Practice, 34(2), 85-102.

10.1016/S0965-8564(98)00065-2
22

Nandi A., Pal A. K. (2021), Interpreting Machine Learning Models: Learn Model Interpretability and Explainability Methods, Apress LP.

23

Park H., Haghani A., Zhang X. (2016) Interpretation of Bayesian Neural Networks for Predicting the Duration of Detected Incidents, J Intell Transp.

10.1080/15472450.2015.1082428
24

Prokhorenkova L., Gusev G., Vorobev A. et al. (2017), CatBoost: Unbiased Boosting with Categorical Features, ArXiv Preprint ArXiv:1706.09516.

25

Tang J., Zheng L., Han C., Liu F., Cai J. (2020), Traffic Incident Clearance Time Prediction and Influencing Factor Analysis using Extreme Gradient Boosting Model, Journal of Advanced Transportation.

10.1155/2020/6401082
26

Uebersax J.S. (1999), Probit Latent Class Analysis with Dichotomous or Ordered Category Measures: Conditional Independence/Dependence Models, Applied Psychological Measurement, 23(4), 283-297.

10.1177/01466219922031400
27

Vacek P.M. (1985), The Effect of Conditional Dependence on the Evaluation of Diagnostic Tests, Biometrics, 41(4), 959-968.

10.2307/25309673830260
28

Wei C. H., Lee Y. (2007), Sequential Forecast of Incident Duration using Artificial Neural Network Models, Accident Analysis & Prevention, 39(5), 944-954.

10.1016/j.aap.2006.12.01717303059
29

Won M., Chang G. L. (2019), Ensemble Model to Estimate Incident Clearance Durations using Sequential Partitioning Process and Robust Regression, Transportation Research Record, 2673(8), 554-566.

10.1177/0361198119842115
30

Wu W. W., Chen S. Y., Zheng C. J. (2011), Traffic Incident Duration Prediction Based on Support Vector Regression, In ICCTP 2011, Towards Sustainable Transportation Systems, 2412-2421.

10.1061/41186(421)24122097838
31

Yang C., Chen M., Yuan Q. (2021), The Application of XGBoost and SHAP to Examining the Factors in Freight Truck-related Crashes: An Exploratory Analysis, Accident Analysis & Prevention, 158, 106153.

10.1016/j.aap.2021.10615334034073
32

Yang Y., Yin Y., Wang Y., Meng R., Yuan Z. (2023), Modeling of Freeway Real-time Traffic Crash Risk Based on Dynamic Traffic Flow Considering Temporal Effect Difference, Journal of Transportation Engineering, Part A: Systems, 149(7), 04023063.

10.1061/JTEPBS.TEENG-7717
33

Yun G. R., Bae S. H. (2021), Analysis of Incident Impact Factors and Development of SMOGN-DNN Model for Prediction of Incident Clearance Time, Journal of the Korea Institute of Intelligent Transportation Systems, 20(4), 46-56.

10.12815/kits.2021.20.4.46
34

Zeng Q., Wang F., Chen T., Sze N. N. (2023), Incorporating Real-time Weather Conditions into Analyzing Clearance Time of Freeway Accidents: A Grouped Random Parameters Hazard-based Duration Model with Time-varying Ccovariates, Analytic Methods in Accident Research, 38, 100267.

10.1016/j.amar.2023.100267
35

Zhao H., Gunardi W., Liu Y., Kiew C., Teng T. H., Yang X. B. (2022), Prediction of Traffic Incident Duration using Clustering-based Ensemble Learning Method, Journal of Transportation Engineering, Part A: Systems, 148(7), 04022044.

10.1061/JTEPBS.0000688
36

Zhao Y., Deng W. (2022), Prediction in Traffic Accident Duration Based on Heterogeneous Ensemble Learning, Applied Artificial Intelligence, 36(1), 2018643.

10.1080/08839514.2021.2018643
37

Zou Y., Han W., Zhang Y., Tang J., Zhong X. (2023), Analyzing Freeway Traffic Incident Clearance Time using a Deep Survival Model, Journal of Transportation Engineering, Part A: Systems, 149(10), 04023101.

10.1061/JTEPBS.TEENG-7653
Information
  • Publisher :Korean Society of Transportation
  • Publisher(Ko) :대한교통학회
  • Journal Title :Journal of Korean Society of Transportation
  • Journal Title(Ko) :대한교통학회지
  • Volume : 42
  • No :3
  • Pages :370-384
  • Received Date : 2024-05-29
  • Revised Date : 2024-06-13
  • Accepted Date : 2024-06-20
  • DOI :https://doi.org/10.7470/jkst.2024.42.3.370
Journal Informaiton Journal of Korean Society of Transportation Journal of Korean Society of Transportation
Archives
: 2015 Vol.34~
  • NRF
  • KOFST
  • crossref cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close