A Novel Host Based Intrusion Detection System using Supervised Learning by Comparing SVM over Random Forest

In order to foresee future developments in host-based intrusion detection systems, this research compares the efficacy of SVM and Random Forest, two supervised-learning-based models. Classification is performed using an SVM algorithm with a sample size of n = 10 and a Random Forest algorithm with a sample size of n = 10, both with a g-power value of 80% and datasets collected from various web sources with recent study findings and threshold 0.05%, confidence interval 95% mean and standard deviation iterated 20 times to obtain data. For the implementation, a further test will be used. According to the data analysis, SVM outperforms Random Forest in terms of accuracy (95.89). (94.12). (p>0.05) In this investigation, no significant differences were found between the groups. Support Vector Machine method outperformed the Random Forest method in the intrusion detection system.