A Novel Method to Analyze a Server Instance's Performance During a Crypto‐Jacking Attack Using Novel Random Forest Algorithm Compared with Logistic Regression
K. Mahesh Reddy, F. Mary Harin Fernandez · 2025
This research focused on identifying cloud server attacks. Regular security audits can assist in locating system flaws and vulnerabilities that hackers might misuse. Audits should be carried out by certified security experts who can offer proposals for improvement. Discovering new hazards and weaknesses can be aided by staying current with security trends. The study is divided into two groups: one using the logistic regression (LR) method with a sample size of ten, and the other using a novel random forest (RF) method with an analogous sample size of ten and determined using ClinCalc software with α = 0.05 and using different training and testing splits, a preliminary test power metric of p=0.8 was found for predicting unauthorized bitcoin cloud server operations during currency transfers. Using Independent samples t-tests (p>0.05), the novel RF model shows improved accuracy at 94.7970%, which is higher than that of LR at 92.9680%. However, the model has a significance value of 0.220. In terms of accuracy, the novel RF outperforms LR.