Performance Evaluation of Multi Layer Perceptron and Support Vector Machine in Email Spam Detection

The focus of this study is to analyze email spam detection using two advanced classification algorithms: the Novel Multilayer Perceptron and the Support Vector Machine (SVM). The research was conducted on a dataset that includes <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{5, 1 7 2}$</tex> emails. Both algorithms were applied to this dataset to evaluate their effectiveness in classifying emails as spam or not. For this study, the sample size for each group was set at 25, calculated using <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$G$</tex> power with a confidence level of 95 %. The goal was to compare the performance of the Novel Multilayer Perceptron against the Support Vector Machine in terms of accuracy and sensitivity. The results showed that the Novel Multilayer Perceptron achieved an accuracy of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{9 1. 6 4 \%}$</tex>, which was higher compared to the Support Vector Machine, which achieved an accuracy of 89.78 %. Additionally, the Novel Multilayer Perceptron had a lower mean error rate than the Support Vector Machine, indicating better performance. The statistical analysis of the results revealed a significant difference with a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$p$</tex>-value of 0.02, which is less than the standard significance level of 0.05. This suggests that the Novel Multilayer Perceptron is more effective than the Support Vector Machine for detecting spam emails.