Maximizing Facial Emotion Recognition Using Artificial Neural Network Algorithm Compared with Novel Linear Regression Algorithm

The purpose of this endeavor is to achieve face recognition in real time through the utilization of machine learning algorithms. Materials and Methods: The speed and precision of real-time facial recognition software play a significant part in optimizing the effectiveness of machine learning and humanoid robots in human-computer interaction. In this study, two algorithms were utilized, and each algorithm underwent 10 iterations, totaling 20 iterations. The sample size was 1588, with a Gpower 80% (=0.05 and power=0.80) setting. The sample was obtained from the Kaggle website. Results: Artificial Neural Network Algorithm (86.92%) classify items and boosts the assessed accuracy over novel linear regression Algorithm (79.43%) by an no significant value of p=0.11 (p>0.05) for both accuracy and loss, which is statistically insignificant, suggesting that the two algorithms don't significantly differ from one another. The loss for significance value of Artificial Neural Network Algorithm is 13.08% and the loss for significance value of Novel Linear Regression Algorithm is 20.57%. Conclusion: Based on the analysis, the Artificial Neural Network (ANN) Algorithm demonstrates superior performance compared to the Novel Linear Regression Algorithm in terms of accuracy. The ANN Algorithm achieved an accuracy rate of 86.92%, outperforming the brand-new linear regression Algorithm, which attained an accuracy rate of 79.43%. Consequently, the findings suggest that the ANN Algorithm is more adept at making accurate predictions in comparison to the Novel Algorithm for linear regression.