Smart decision making system to estimate the level of threats in areas involved using the machine learning algorithms

Maryam Rezaee¹

IOI Code : ABEB-ABEBAB
Publication : 4th International Conference on Recent Advances in Engineering, Innovation and Technology - Belgium(eitconf.com)
Abstract :
    The main issue in this research is to assess the level of insecurity of an area based on a set of operational and environmental characteristics with the aim of supporting intelligent decision-making in high-risk situations. For this purpose, machine learning algorithms including artificial neural networks, neo-biz, support vector machines, and nearest neighbor algorithms were used. The data required for the research were generated with a simulation and labeling approach and, after performing preprocessing steps, were used to train and evaluate the models.The performance of the algorithms was evaluated based on common evaluation and classification criteria. The results showed that the artificial neural network, by achieving the highest accuracy and more appropriate resolution, has a better performance in identifying different levels of insecurity than other algorithms. The NioBiz algorithm showed an acceptable performance, while the support vector machine and nearest neighbor had a weaker performance in terms of accuracy and detection capability. Overall, the findings of this study indicate the appropriate efficiency of machine learning methods, especially neural networks, in analyzing complex situations and supporting intelligent decision-making in operational environments and the battlefield.

Keywords : machine learning algorithm battlefield decision support system high risk environments neural networks