Convolutional Neural Network vs. Random Forest for Accurate Human Face Recognition in Cybercrime Analysis

In the context of cybercrime analysis, this study examines the efficacy of several facial recognition systems, with a particular emphasis on two techniques: Random Forest and Convolutional neural network models (The CNN network). There were two groups in the investigation, each with ten samples. An alpha value of 0.05, a beta value of 0.2, and a power value of 0.8 were the three main analysis parameters. According to the findings, the CNN approach had an accuracy of 74.7 %, while the Random Forest technique had 77.7 %. The statistical importance of the difference in accuracy between the two approaches is demonstrated by the p-value of 0.003, which is below the significance level of 0.05. This indicates that Random Forest performs far better than CNN in both facial recognition and human emotions classification..