Predicting Hurricane Evacuation Decisions with Interpretable Machine Learning Methods

Abstract

Facing the escalating effects of climate change, it is critical to improve the prediction and understanding of the hurricane evacuation decisions made by households in order to enhance emergency management. Current studies in this area often have relied on psychology-driven linear models, which frequently exhibited limitations in practice. The present study proposed a novel interpretable machine learning approach to predict household-level evacuation decisions by leveraging easily accessible demographic and resource-related predictors, compared to existing models that mainly rely on psychological factors. An enhanced logistic regression model (that is, an interpretable machine learning approach) was developed for accurate predictions by automatically accounting for nonlinearities and interactions (that is, univariate and bivariate threshold effects). Specifically, nonlinearity and interaction detection were enabled by low-depth decision trees, which offer transparent model structure and robustness. A survey dataset collected in the aftermath of Hurricanes Katrina and Rita, two of the most intense tropical storms of the last two decades, was employed to test the new methodology. The findings show that, when predicting the households’ evacuation decisions, the enhanced logistic regression model outperformed previous linear models in terms of both model fit and predictive capability. This outcome suggests that our proposed methodology could provide a new tool and framework for emergency management authorities to improve the prediction of evacuation traffic demands in a timely and accurate manner.

Keywords

Artificial Intelligence (AI) / Decision-making modeling / Hurricane evacuation / Interpretable machine learning / Nonlinearity and interaction detection

References

Ahmed, M.A., A.M. Sadri, and M. Hadi. 2020. Modeling social network influence on hurricane evacuation decision consistency and sharing capacity. Transportation Research Interdisciplinary Perspectives 7: Article 100180.

Alawadi, R., P. Murray-Tuite, D. Marasco, S. Ukkusuri, and Y. Ge. 2020. Determinants of full and partial household evacuation decision making in Hurricane Matthew. Transportation Research Part D: Transport and Environment 83: Article 102313.

Anyidoho, P.K., X. Ju, R.A. Davidson, and L.K. Nozick. 2023. A machine learning approach for predicting hurricane evacuee destination location using smartphone location data. Computational Urban Science 3(1): Article 30.

Bagloee SA, Johansson KH, Asadi M. A hybrid machine-learning and optimization method for contraflow design in post-disaster cases and traffic management scenarios. Expert Systems with Applications, 2019, 124: 67-81

Baker EJ. Hurricane evacuation behavior. International Journal of Mass Emergencies & Disasters, 1991, 9(2): 287-310

Bhavan, A., P. Chauhan, and R.R. Shah. 2019. Bagged support vector machines for emotion recognition from speech. Knowledge-Based Systems 184: Article 104886.

Burris, J.W., R. Shrestha, B. Gautam, and B. Bista. 2015. Machine learning for the activation of contraflows during hurricane evacuation. In Proceedings of the 2015 IEEE global humanitarian technology conference (GHTC), 8–11 October 2015, Seattle, WA, USA, 254–258.

Cahyanto I, Pennington-Gray L, Thapa B, Srinivasan S, Villegas J, Matyas C, Kiousis S. An empirical evaluation of the determinants of tourist’s hurricane evacuation decision making. Journal of Destination Marketing & Management, 2014, 2(4): 253-265

Census Bureau, U.S. 2019. Coastline America. http://www.ncdc.noaa.gov/billions. Accessed 8 Dec 2023.

Dash N, Gladwin H. Evacuation decision making and behavioral responses: Individual and household. Natural Hazards Review, 2007, 8(3): 69-77

Davidson RA, Nozick LK, Wachtendorf T, Blanton B, Colle B, Kolar RL, DeYoung S, Dresback KM An integrated scenario ensemble-based framework for hurricane evacuation modeling: Part 1 – Decision support system. Risk Analysis, 2020, 40(1): 97-116

Dillman DA. Mail and Internet surveys: The tailored design method – 2007 update with new Internet, visual, and mixed-mode guide, 2011, Hoboken, NJ: Wiley

Dumitrescu E, Hué S, Hurlin C, Tokpavi S. Machine learning for credit scoring: Improving logistic regression with non-linear decision-tree effects. European Journal of Operational Research, 2022, 297(3): 1178-1192

Ersing, R.L., C. Pearce, J. Collins, M.E. Saunders, and A. Polen. 2020. Geophysical and social influences on evacuation decision-making: The case of hurricane Irma. Atmosphere 11(8): Article 851.

Fraser T. Fleeing the unsustainable city: Soft policy and the dual effect of social capital in hurricane evacuation. Sustainability Science, 2022, 17(5): 1995-2011

Gladwin H. Gillis Peacock W, Gladwin H, Hearn Morrow B. Warning and evacuation: A night for hard houses. Hurricane Andrew: Ethnicity, gender and the sociology of disasters, 1997, London: Taylor & Francis 52-74.

Gladwin H, Lazo JK, Morrow BH, Peacock WG, Willoughby HE. Social science research needs for the hurricane forecast and warning system. Natural Hazards Review, 2007, 8(3): 87-95

Hasan S, Ukkusuri S, Gladwin H, Murray-Tuite P. Behavioral model to understand household-level hurricane evacuation decision making. Journal of Transportation Engineering, 2011, 137(5): 341-348

Horney JA, MacDonald PDM, Van Willigen M, Berke PR, Kaufman JS. Factors associated with evacuation from Hurricane Isabel in North Carolina, 2003. International Journal of Mass Emergencies & Disasters, 2010, 28(1): 33-58

Huang S-K, Lindell MK, Prater CS, Wu H-C, Siebeneck LK. Household evacuation decision making in response to Hurricane Ike. Natural Hazards Review, 2012, 13(4): 283-296

Huang S-K, Lindell MK, Prater CS. Who leaves and who stays? A review and statistical meta-analysis of hurricane evacuation studies. Environment and Behavior, 2016, 48(8): 991-1029

Huang, S.-K., M.K. Lindell, and C.S. Prater. 2017. Multistage model of hurricane evacuation decision: Empirical study of Hurricanes Katrina and Rita. Natural Hazards Review 18(3): Article 05016008.

Kildow, J.T., C.S. Colgan, and J. Pat. 2016. State of the U.S. ocean and coastal economies 2016 update. https://cbe.miis.edu/noep_publications/18. Accessed 30 Dec 2023.

Kim, P.B. 2015. Interactive and interpretable machine learning models for human machine collaboration. Ph.D. thesis. Massachusetts Institute of Technology, Cambridge, MA, USA.

Kim H, Park J-H, Lee D, Yang Y-S. Establishing the methodologies for human evacuation simulation in marine accidents. Computers & Industrial Engineering, 2004, 46(4): 725-740

Knabb, R.D., D.P. Brown, and J.R. Rhome. 2006. Tropical cyclone report: Hurricane Rita 18–26 September 2005. https://www.nhc.noaa.gov/data/tcr/AL182005_Rita.pdf. Accessed 30 Dec 2023.

Knabb, R.D., J.R. Rhome, and D.P. Brown. 2023. Tropical cyclone report: Hurricane Katrina, 23–30 August 2005. https://www.nhc.noaa.gov/data/tcr/AL122005_Katrina.pdf. Accessed 30 Dec 2023.

Lazo JK, Bostrom A, Morss RE, Demuth JL, Lazrus H. Factors affecting hurricane evacuation intentions. Risk Analysis, 2015, 35(10): 1837-1857

Lee D, Park J-H, Kim H. A study on experiment of human behavior for evacuation simulation. Ocean Engineering, 2004, 31(8–9): 931-941

Li D, Zhang Z, Alizadeh B, Zhang Z, Duffield N, Meyer MA, Thompson CM, Gao H, Behzadan AH. A reinforcement learning-based routing algorithm for large street networks. International Journal of Geographical Information Science, 2023

Lindell MK. Badiru AB, Racz L. Evacuation planning, analysis, and management. Handbook of emergency response: A human factors and systems engineering approach, 2013, Boca Raton, FL: CRC Press 121-149.

Lindell MK. Rodríguez H, Donner W, Trainor JE. Communicating imminent risk. Handbook of disaster research, 2018, Cham: Springer 449-477

Lindell MK. The Routledge handbook of urban disaster resilience: Integrating mitigation, preparedness, and recovery planning, 2019, London: Routledge

Lindell MK, Perry RW. The protective action decision model: Theoretical modifications and additional evidence. Risk Analysis, 2012, 32(4): 616-632

Lindell MK, Lu J-C, Prater CS. Household decision making and evacuation in response to Hurricane Lili. Natural Hazards Review, 2005, 6(4): 171-179

Ling L, Ukkusuri SV, Murray-Tuite P, Lee S, Ge Y. Modeling the causality of received information, certainty, and decision making in hurricane evacuations. SSRN Electronic Journal, 2022

Lo SM, Liu M, Zhang PH, Yuen RK. An artificial neural-network based predictive model for pre-evacuation human response in domestic building fire. Fire Technology, 2009, 45: 431-449

Metaxa-Kakavouli, D., P. Maas, and D.P. Aldrich. 2018. How social ties influence hurricane evacuation behavior. In Proceedings of the ACM on Human-Computer Interaction 2(CSCW): Article 122.

Meyer MA, Mitchell B, Purdum JC, Breen K, Iles RL. Previous hurricane evacuation decisions and future evacuation intentions among residents of southeast Louisiana. International Journal of Disaster Risk Reduction, 2018, 31: 1231-1244

Molnar, C. 2020. Interpretable machine learning. https://christophm.github.io/interpretable-ml-book/. Accessed 5 Feb 2024.

Morss RE, Demuth JL, Lazo JK, Dickinson K, Lazrus H, Morrow BH. Understanding public hurricane evacuation decisions and responses to forecast and warning messages. Weather and Forecasting, 2016, 31(2): 395-417

Nara A, Machiani SG, Luo N, Ahmadi A, Robinett K, Tominaga K, Ahmadi A, Robinett K Nara A, Tsou M-H Learning dependence relationships of evacuation decision making factors from Tweets. Empowering human dynamics research with social media and geospatial data analytics, 2021, Cham: Springer 113-138

Petty RE, Cacioppo JT. The elaboration likelihood model of persuasion, 1986, Cham: Springer

Ramaguru, G.P., and V.D.K. Pasupuleti. 2019. Human evacuation simulation during disaster: A web tool. In Innovative research in transportation infrastructure: Proceedings of ICIIF 2018, ed. D. Deb, V.E. Balas, R. Dey, and J. Shah, 33–43. Singapore: Springer.

Roy, K.C., S. Hasan, A. Culotta, and N. Eluru. 2021. Predicting traffic demand during hurricane evacuation using real-time data from transportation systems and social media. Transportation Research Part C: Emerging Technologies 131: Article 103339.

Rudin C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 2019, 1(5): 206-215

Sadri, A.M., S.V. Ukkusuri, and H. Gladwin. 2017. The role of social networks and information sources on hurricane evacuation decision making. Natural Hazards Review 18(3): Article 04017005.

Sarwar MT, Anastasopoulos PCh, Ukkusuri SV, Murray-Tuite P, Mannering FL. A statistical analysis of the dynamics of household hurricane-evacuation decisions. Transportation, 2018, 45: 51-70

Schorr, J.P. 2015. Bridging the gap between social and transportation networks: An integrated, dynamic evacuation decision making model. https://scholarspace.library.gwu.edu/concern/gw_etds/c534fn99d. Accessed 30 Dec 2023.

Sorensen JH. Hazard warning systems: Review of 20 years of progress. Natural Hazards Review, 2000, 1(2): 119-125

Sun W, Bocchini P, Davison BD. Applications of artificial intelligence for disaster management. Natural Hazards, 2020, 103(3): 2631-2689

Tanim, S.H., B.M. Wiernik, S. Reader, and Y. Hu. 2022. Predictors of hurricane evacuation decisions: A meta-analysis. Journal of Environmental Psychology 79: Article 101742.

Vreugdenhil, B.J., N. Bellomo, and P.S. Townsend. 2015. Using crowd modelling in evacuation decision making. In Proceedings of the ISCRAM 2015 conference. https://idl.iscram.org/files/bjvreugdenhil/2015/1288_B.J.Vreugdenhil_etal2015.pdf. Accessed 6 Feb 2024.

Xu N, Lovreglio R, Kuligowski ED, Cova TJ, Nilsson D, Zhao X. Predicting and assessing wildfire evacuation decision-making using machine learning: Findings from the 2019 Kincade Fire. Fire Technology, 2023, 59(2): 793-825

Yang, K., R.A. Davidson, B. Blanton, B. Colle, K. Dresback, R. Kolar, L.K. Nozick, J. Trivedi, and T. Wachtendorf. 2019. Hurricane evacuations in the face of uncertainty: Use of integrated models to support robust, adaptive, and repeated decision-making. International Journal of Disaster Risk Reduction 36: Article 101093.

Yang, J., R. Shi, and B. Ni. 2021. MedMNIST classification decathlon: A lightweight AutoML benchmark for medical image analysis. In Proceedings of the 2021 IEEE 18th International symposium on biomedical imaging (ISBI), 13–16 April 2021, Nice, France, 191–195.

Zhao X, Lovreglio R, Kuligowski E, Nilsson D. Using artificial intelligence for safe and effective wildfire evacuations. Fire Technology, 2021, 57: 483-485

Zhao, X., R. Lovreglio, and D. Nilsson. 2020. Modelling and interpreting pre-evacuation decision-making using machine learning. Automation in Construction 113: Article 103140.

Zhu, R., B. Becerik-Gerber, J. Lin, and N. Li. 2023. Behavioral, data-driven, agent-based evacuation simulation for building safety design using machine learning and discrete choice models. Advanced Engineering Informatics 55: Article 101827.