Comparative Analysis of Cyber Crime Breaches Using Random Forest Over Decision Tree to Improve Accuracy
K. Venkateshwarprasad, G. Arul Freeda Vinodhini, Vijit Joon, V. Mathivanan · 2024
Random Forest is preferred over Decision Tree in a comparative examination of cybercrime breaches in order to increase accuracy. A unique innovative RF-Classifier is employed to attain accuracy. A total of 20 samples are taken for the Random Algorithm ($\mathbf{N} \boldsymbol{=} \mathbf{1 0}$) and Decision Tree Algorithm ($\mathbf{N} \boldsymbol{=}$10) calculations, which are done utilizing two groups. The Kaggle Library's cyber breaches dataset is used to test accuracy and loss. The Random Forest method yielded a 99.39% success rate and a 1.61% loss compared to the Decision Tree algorithm's 91.60% success rate and 9.49% loss. Ultimately, the significance value obtained by the Random Forest and Decision Tree algorithms with Independent Samples T-test is$\mathbf{p}=\mathbf{0. 3 7 1}(p>0.05)$, indicating that two groups are statistically not significant.