Application of a logistic regression model in decision-making on determining the number of forces involved in the elimination of forest fires

Introduction. Forecasting the number of involved forces for elimination and localization of forest fires is an important and urgent task that affects the efficiency of the work carried out. However, the use of traditional methods of statistical forecasting does not allow to obtain a reliable assessment of the target indicator, due to the lack of a number of features in the analysis, the consequence of which is a decrease in the effectiveness of decisions.

Objective. Investigation of the possibility of applying the logistic regression model to make decisions on the number of forces to be used for localization and suppression of forest fires at the initial stage of the fire.

Research methods. The application of the logistic regression method was evaluated on the basis of a database of forest fires in the territory of the Leningrad region in the period from 2015 to 2023, in which 16 features were identified. The logistic regression model allows training on data with different types of distribution, including binomial, Poisson, Bernoulli and other types of distribution. The mathematical apparatus used in the model allows us to estimate the posterior probabilities for assigning training objects to the appropriate classes.

Results. The results of the model training evaluation in the form of error matrices and classification reports are presented as results, and visualization of the decision boundaries for the cases of using two and three features is performed. The results show that the best accuracy was achieved using all available features.

Conclusion. The research of forest fires data in the Leningrad region has shown that there are factors that are not taken into account when making plans for the involvement of forces and resources. The use of machine learning models and, in particular, logistic regression, proposed in this study, can improve the validity and efficiency of decisions to determine the number of forces to be involved in forest fires.

1. Vilkov V.B., Gorshkova E.E., Chernykh A.K. Solving the problem of finding the optimal route for patrolling active forest fires in a given area. Scientific and analytical journal “Bulletin of the St. Petersburg University of the State Fire Service of the Ministry of Emergency Situations of Russia”. 2021; 3:90-98. EDN ZYOYFF. (rus).

2. Molchanov A.V. Conceptual model of formation of rational structure of group of forces of various departmental accessory, at elimination of emergency situations in the woods of regional character which arose owing to forest fires. Scientific and Educational Problems of Civil Protection. 2015; 4(27):95-102. EDN VCITFJ. (rus).

3. Karapuzikov A.A., Dyakov V.F., Koksharov A.V., Dyakov M.V., Stavrinidi S.Y., Belkin D.S. On the issue of managing forces and means for extinguishing forest fires. Technospheric Safety. 2020; 2(27):16-27. EDN URZSEG. (rus).

4. Matveev A.V., Bogdanova E.M. The classification for the methods of prediction emergency situations. National Security and Strategic Planning. 2018; 4(24):61-70. EDN YTPZYL. (rus).

5. Stankevich T.S. Forecasting the spatial behavior of a forest fire at uncertainty and instability of the process. News of Higher Educational Institutions. Forest Magazine. 2021; 1(379):20-34. DOI: 10.37482/0536-1036-2021-1-20-34. (rus).

6. Stankevich T.S. Development of operational prediction method of forest fire dynamics based on artificial intelligence and deep machine learning. Bulletin of the Irkutsk State Technical University. 2018; 22(9):111-120. DOI: 10.21285/­1814-3520-2018-9-111-120 (rus).

7. Matveev A.V., Matiev R.T. Decision-making during fires in mountainous areas : a comparative analysis of monitoring methods. National Security and Strategic Planning. 2023; 2(42):76-90. DOI: 10.37468/2307-1400-2023-2-76-90 (rus).

8. Madaio M., Chen S.T., Haimson O., Zhang W., Cheng X., Hinds-Aldrich M. et al. Firebird: Predicting fire risk and prioritizing fire inspections in Atlanta. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2016; 185-194. DOI: 10.1145/2939672.2939682

9. Nikolopoulos E.I., Destro E., Bhuiyan M., Borga M., Anagnostou E. Evaluation of predictive models for post-fire debris flow occurrence in the western United States. Natural Hazards and Earth System Sciences. 2018; 18(9):2331-2343. DOI: 10.5194/nhess-18-2331-2018

10. Pham B.T., Jaafari A., Avand M., Al-Ansari N., Dinh Du T., Yen H.P. et al. Performance evaluation of machine learning methods for forest fire modeling and prediction. Symmetry. 2020; 12(6):1022. DOI: 10.3390/sym12061022

11. Maximov A.V. Decision support methods in emergency management: a review of research. National Security and Strategic Planning. 2023; 2(42):91-102. DOI: 10.37468/2307-1400-2023-2-91-102 (rus).

12. Butyrsky E.Yu., Matveev A.V. Mathematical modeling of systems and processes. 2022; 733. DOI 10.37468/book_011222 (rus).

13. Ray S. A quick review of machine learning algorithms. 2019 International conference on machine learning, big data, cloud and parallel computing (COMITCon). 2019; 35-39. DOI: 10.1109/COMITCon.2019.8862451

14. Abid F. A survey of machine learning algorithms based forest fires prediction and detection systems. Fire Technology. 2021; 57(2):559-590. DOI: 10.1007/s10694-020-01056-z

15. Sayad Y.O., Mousannif H., Al Moatassime H. Predictive modeling of wildfires: A new dataset and machine learning approach. Fire Safety Journal. 2019; 104:130-146. DOI: 10.1016/j.firesaf.2019.01.006

16. Yang Y., Loog M. A benchmark and comparison of active learning for logistic regression. Pattern Recognition. 2018; 83:401-415. DOI: 10.1016/j.patcog.2018.06.004

17. Medvedev D.V. Implementation of the ANFIS neuro-fuzzy system for forest fire management. Scientific and Analytical Journal “Bulletin of the St. Petersburg University of the State Fire Service of the Ministry of Emergency Situations of Russia”. 2023; 4:185-198. DOI: 10.61260/2218-130X-2024-2023-4-185-198 (rus).

18. Karthiga R., Usha G., Raju N., Narasimhan K. Transfer learning based breast cancer classification using one-hot encoding technique. 2021 International Conference on Artificial Intelligence and Smart Systems (ICAIS). 2021; 115-120. DOI: 10.1109/ICAIS50930.2021.9395930

19. Salehi F., Abbasi E., Hassibi B. The impact of regularization on high-dimensional logistic regression. Advances in Neural Information Processing Systems. 2019; 32.

20. Liang X., Jacobucci R. Regularized structural equation modeling to detect measurement bias: Evaluation of lasso, adaptive lasso, and elastic net. Structural Equation Modeling : а Multidisciplinary Journal. 2020; 27(5):722-734. DOI: 10.1080/10705511.2019.1693273

21. Molovtsev M.D., Sineva I.S. Classification algorithms analysis in the forest fire detection problem. 2019 International Conference “Quality Management, Transport and Information Security, Information Technologies” (IT&QM&IS). IEEE, 2019; 548-553. DOI: 10.1109/ITQMIS.2019.8928398

22. Hayaty M., Muthmainah S., Ghufran S.M. Random and synthetic over-sampling approach to resolve data imbalance in classification. International Journal of Artificial Intelligence Research. 2020; 4(2):86-94. DOI: 10.29099/ijair.v4i2.152