The 2017 Puebla, Mexico, earthquake event led to significant damage in many buildings in Mexico City. In the months following the earthquake, civil engineering students conducted detailed building assessments throughout the city. They collected building damage information and structural characteristics for 340 buildings in the Mexico City urban area, with an emphasis on the Roma and Condesa neighborhoods where they assessed 237 buildings. These neighborhoods are of particular interest due to the availability of seismic records captured by nearby recording stations, and preexisting information from when the neighborhoods were affected by the 1985 Michoacán earthquake. This article presents a case study on developing a damage prediction model using machine learning. It details a framework suitable for working with future post-earthquake observation data. Four algorithms able to perform classification tasks were trialed. Random forest, the best performing algorithm, achieves more than 65% prediction accuracy. The study of the feature importance for the random forest shows that the building location, seismic demand, and building height are the parameters that influence the model output the most.
AllenLBrzevSCharlesonAWScawthornCSilvaV (2015) GEM building taxonomy—An open global building classification system. In: Proceedings of 2015 New Zealand society for earthquake engineering annual technical conference, Rotorua, New Zealand, pp. 1–8. Available at: http://www.nzsee.org.nz/db/2015/Papers/O-50_Allen.pdf (accessed 15 June 2020).
Arellano-MendezEJuarez-GarciaHGomez-BernalA (2004) Vulnerabilidad Sismica de la Colonia Roma, Ciudad de Mexico. In: XIV Congreso Nacional de Ingeniería Estructural, pp. 1–23. Acapulco, Mexico: Sociedad Mexicana de Ingeniería Estructural. Available at: http://www.smie.org.mx/SMIE_Articulos/co/co_13/te_05/ar_01.pdf (accessed 15 June 2020)
5.
AslaniHMirandaE (2005) Fragility assessment of slab-column connections in existing non-ductile reinforced concrete buildings. Journal of Earthquake Engineering9(6): 777–804.
Colegio de Ingenieros Civiles de México (CICM) (2017) Resumen preliminar de danos de los inmeubles inspeccionados por las brigadas del CICM del sismo del 19/09/2017. Available at: https://www.sismosmexico.org/informes
EllingwoodBRCelikOCKinaliK (2007) Fragility assessment of building structural systems in Mid-America. Earthquake Engineering & Structural Dynamics36(13): 1935–1952.
Gomez-BernalASaragoniR (2002) Respuesta Dinámica en suelos estratificados durante terremotos. In: VIII Jornadas Chilenas de Sismología e Ing Antisísmica (CD Rom, Paper No. 111), Valparaiso, Chile, 25–27 April.
GuyonIElisseeffA (2003) An introduction to variable and feature selection. Journal of Machine Learning Research3: 1157–1182.
22.
HastieTTibshiraniRFriedmanJ (2009) The Elements of Statistical Learning. New York: Springer.
23.
HoneggerM (2018) Shedding Light on Black Box Machine Learning Algorithms: Development of an Axiomatic Framework to Assess the Quality of Methods that Explain Individual Predictions. Karlsruhe Institute of Technology. Available at: http://arxiv.org/abs/1808.05054
24.
HutchinsonTCRay ChaudhuriS (2006) Simplified expression for seismic fragility estimation of sliding-dominated equipment and contents. Earthquake Spectra22(3): 709–732.
JamesGWittenDHastieTTibshiraniR (2013) An Introduction to Statistical Learning: With Applications in R. New York: Springer.
27.
KianiJCampCPezeshkS (2019) On the application of machine learning techniques to derive seismic fragility curves. Computers & Structures218: 108–122.
28.
KovačevićMStojadinovićZMarinkovićDStojadinovićB (2018) Sampling and machine learning methods for a rapid earthquake loss assessment system. In: 11th U.S. national conference on earthquake engineering, Los Angeles, CA, 25–29 June.
LeeK-F (2018) AI Superpowers: China, Silicon Valley, and the New World Order. Boston, MA: Houghton Mifflin Harcourt.
31.
LundbergSMLeeS-I (2017) A unified approach to interpreting model predictions. In: GarnettIGLuxburgUVBengioSWallachHFergusRVishwanathanSGarnettR (eds) Advances in Neural Information Processing Systems 30. New York: Curran Associates, Inc., pp. 4765–4774.
32.
MangalathuS.BurtonH. V. (2019). Deep learning-based classification of earthquake-impacted buildings using textual damage descriptions. International Journal of Disaster Risk Reduction, 36 (February). https://doi.org/10.1016/j.ijdrr.2019.101111
33.
MangalathuS.SunH.NwekeC. C.YiZ.BurtonH. V. (2020). Classifying earthquake damage to buildings using machine learning. Earthquake Spectra, 36(1), 183–208. https://doi.org/10.1177/8755293019878137
34.
MayoralJMAsimakiDTepalcapaSWoodCRoman-delaSanchaAHutchinsonTFrankeKMontalvaG (2019) Site effects in Mexico City basin: Past and present. Soil Dynamics and Earthquake Engineering121: 369–382.
MayoralJ. M.RomanA.De La RosaD.AlcarazM.RivasR. (2019). Key findings and observations following the September 19th, 2017 Mw 7.1 Puebla-Mexico City earthquake. In SilvestriF.MoraciN. (Eds.), Earthquake Geotechnical Engineering for Protection and Development of Environment and Constructions (pp. 833–844). CRC Press. Available at: https://doi.org/10.1201/9780429031274
37.
MillerT (2019) Explanation in artificial intelligence: Insights from the social sciences. Artificial Intelligence267: 1–38.
38.
MitchellTM (1997) Machine Learning. New York: McGraw-Hill.
PorterKKiremidjianA (2001) Assembly-Based Vulnerability of Buildings and Its Uses in Seismic Performance Evaluation and Risk Management Decision-Making. John A. Blume Earthquake Engineering Center. Available at: http://purl.stanford.edu/qf102hx9901
Python Software Foundation (2019) Python Language Reference (Version 3.8). Available at: http://www.python.org
47.
RibeiroMTSinghSGuestrinC (2016a) Model-agnostic interpretability of machine learning. In: 2016 ICML workshop on human interpretability in machine. Available at: http://arxiv.org/abs/1606.05386
48.
RibeiroMTSinghSGuestrinC (2016b) “Why should I trust you?” Explaining the predictions of any classifier. In: Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. ACM, pp. 1135–1144. Available at: https://www.kdd.org/kdd2016/papers/files/rfp0573-ribeiroA.pdf (accessed 15 June 2020).
49.
RibeiroMTSinghSGuestrinC (2018) Anchors: High-precision model-agnostic explanations. In: Proceedings of the 32nd AAAI conference on artificial intelligence (AAAI’18), pp. 1527–1535. Available at: www.aaai.org
50.
RoeslinSElwoodKJJuarez-GarciaHGomez-BernalADhakalRP (2018a) The September 19th, 2017 Puebla, Mexico Earthquake. In: Proceedings of the New Zealand society for earthquake engineering annual technical conference 2018. Auckland, New Zealand, 15 April, pp. 1–10. Available at:https://researchspace.auckland.ac.nz/handle/2292/44962 (accessed 15 June 2020).
51.
RoeslinSMaQTMGarcíaHJ (2018b) Damage assessment on buildings following the 19th September 2017 Puebla, Mexico Earthquake. Frontiers in Built Environment4: 18.
52.
Román-DeLaSanchaAMayoralJMHutchinsonTCCandiaGMontgomeryJTepalcapaS (2019) Assessment of fragility models based on the Sept 19th, 2017 earthquake observed damage. Soil Dynamics and Earthquake Engineering125: 105707.
53.
RothAE (2005) The Shapley Value. Cambridge: Cambridge University Press.
TangJAlelyaniSLiuH (2014) Feature selection for classification: A review. In: AggarwalCC (ed.) Data Classification: Algorithms and Applications. Boca Raton, FL: CRC Press, pp. 37–64.
Villar-VegaMSilvaV (2017) Assessment of earthquake damage considering the characteristics of past events in South America. Soil Dynamics and Earthquake Engineering99: 86–96.
