Machine Learning Techniques in Structural Fire Risk Prediction
AUTHORS
Jaesung Chang,Dept. Industrial and Information Systems Eng, Soongsil Univ., Sangdoro 369, Dongjakku, Seoul, Republic of Korea
Jaeyoung Yoon,Dept. Industrial and Information Systems Eng, Soongsil Univ., Sangdoro 369, Dongjakku, Seoul, Republic of Korea
Gunho Lee*,Dept. Industrial and Information Systems Eng, Soongsil Univ., Sangdoro 369, Dongjakku, Seoul, Republic of Korea
ABSTRACT
Fires often occur, and the damage caused by them is often irreversible. Fire-prone environments can be identified through historical data, and predictive models are recommended to prevent fires in advance. This study uses a variety of machine learning techniques to build fire prediction models and perform a comparative analysis to predict fires. We use data from local fire departments in South Korea to build fire prediction models using decision trees, random forest, XGBoost, extra tree classification, artificial neural networks, and more. Before creating the fire prediction models, we analyze and significant predictive features of a structural fire. We compared the fire prediction models and showed accuracy, F1-score, precision, and recall. The prediction model built with the random forest is the most accurate, but there is a little difference in the accuracy of each model trained with the extra tree classifier, XGBoost, and neural network. For the F1 Score, the model with a neural network shows the best value.
KEYWORDS
Machine learning methods, Fire prediction, Comparative analysis
REFERENCES
[1] M. S. Won, K. S. Koo, and M. B. Lee, “An analysis of forest fire occurrence hazards by changing temperature and humidity of ten-day intervals for 30 years in spring,” Korean Journal of Agricultural and Forest Meteorology, vol.8, no.4, pp.250-259, (2006)
[2] http://www.nfa.go.kr/nfa/releaseinformation/statisticalinformation/main/=view&cntId=20&category=&pageIdx=&searchCondition=&searchKeyword, March 5, (2020)
[3] S. L. Manzello, R. Blanchi, M. J. Gollner, D. Gorham, S. McAllister, E. Pastor, E. Planas, P. Reszka, and S. Suzuki, “Summary of workshop large outdoor fires and the built environment,” Fire Safety Journal, vol.100, pp.76-92, (2018)
[4] J. Xin, and C. Huang, “Fire risk analysis of residential buildings based on scenario clusters and its application in fire risk management,” Fire Safety Journal, vol.62, Part A, pp.72-78, (2013)
[5] J. Yang and Y. Chen, “Research and application of fire risk assessment system for marketplace buildings,” Procedia Engineering, vol.71, pp.476-480, (2014)
[6] M. Madaio, O. L. Haimson, W. Zhang, X. Cheng, M. Hinds-Aldrich, B. Dilkina, and D.H.P. Chau, “Identifying and prioritizing fire inspections: a case study of predicting fire risk in Atlanta,” Bloomberg Data for Good Exchange, New York, NY, USA. (2015)
[7] M. A. Madaio, “Predictive modeling of building fire risk-designing and evaluating predictive models of fire risk to prioritize property fire inspections,” A metro 21 Research Project, Carnegie Mellon University,13-27, (2018)
[8] O. Jungbluth, U.S. Patent No. 4,196,558. Washington, DC: U.S. Patent and Trademark Office, (1980)
[9] M. Ahrens, “Smoking and fire,” American Journal of Public Health, vol.94, no.7, pp.1076-1077, (2004)
[10] S. Lu, G. Li, P. Mei, and H. Zhang, “Suppressive effects of fire prevention campaign in China: A time series analysis,” Safety Science, vol.86, pp.69-77, (2016)
[11] R. Caruana and A. Niculescu-Mizil, “An empirical comparison of supervised learning algorithms,” In Proceedings of the 23rd international conference on Machine learning, June: 25-29; Pittsburgh, USA, (2006)
[12] S. R. Safavian and D. Landgrebe, “A survey of decision tree classifier methodology,” IEEE transactions on Systems, Man, and Cybernetics, vol.21, no.3, pp.660-674, (1991)
[13] L. Breiman, “Random forests,” Machine Learning, vol.45, no.1, pp.5-32, (2001)
[14] https://towardsdatascience.com/an-intuitive-explanation-of-random-forest-and-extra-trees-classifiers 8507ac21d54b March 7, (2020)
[15] T. Chen and C. Guestrin, “Xgboost: A scalable tree boosting system,” In Proceedings of the 22nd ACM sigkdd international conference on knowledge discovery and data mining, San Francisco, USA, August: 22-27, (2016)
[16] L. Yang, C. W. Dawson, M. R. Brown, and M. Gell, “Neural network and GA approaches for dwelling fire occurrence prediction,” Knowledge-Based Systems, vol.19, no.4, pp.213-219, (2006)
[17] T. Windeatt, “Ensemble MLP classifier design,” In Computational Intelligence Paradigms, Springer, (2008)
[18] https://en.wikipedia.org/wiki/F1_score, March 6, (2020)