IOT Attack Type Detection and Classification Based on Multi-strategy Hybrid and Improved Sparrow Search Algorithm to Optimize Random Forest Model
Qi Hang Wu · 2024
In this paper, the classical random forest model is used as the basis, and the sparrow search algorithm with multi-strategy hybrid is combined to optimize the attack detection and attack type prediction in the Internet of Things system. The aim of the experiment is to improve the performance of the random forest model by using various parameters of the Internet of Things system, so as to realize effective identification and classification of various attacks in the Internet of Things environment. During the experiment, the dataset contained three different types of IOT attacks. After 15 training iterations, the model's prediction results of attack types gradually approached the actual situation, and the initial loss value gradually decreased from 3.2 to 0 in the 14th iteration, which indicates that the model's prediction ability and classification effect have been significantly improved. By analyzing the confusion matrix of the training set, it can be seen that the model achieves 100% accurate prediction for all three types of IOT attacks. In the test set, a total of 270 attack types were predicted, of which 268 were correct and only two were wrong, resulting in a prediction accuracy of 99.26%. This result is similar to the performance of the training set, indicating that the model not only has a good prediction effect, but also shows a strong generalization ability. In summary, this study successfully applied the multi-strategy hybrid improved Sparrow search algorithm to the random forest model, effectively improving the performance of attack detection and classification in IOT systems. This achievement provides important support for ensuring the security of the Internet of Things environment, not only enhances the response ability to potential threats, but also provides new ideas and methods for future related research.