Precision in Classification: A Comparative Study of Logistic Regression, Naive Bayes, LSTM, and CNN for Spam Email Detection

Abstract

This research presents a comprehensive comparative study on the precision of classification algorithms applied to spam email detection. The study focuses on four distinct algorithms: Logistic Regression, Naive Bayes, Long Short-Term Memory (LSTM), and Convolutional Neural Network (CNN). The evaluation encompasses a common dataset to ensure a fair and rigorous assessment of each algorithm’s performance. We investigate and analyze precision metrics, considering the trade- offs between true positives and false positives. The comparative study provides insights into the strengths and limitations of each algorithm in effectively identifying spam emails. The results contribute valuable knowledge to the field of email security and classification, guiding practitioners and researchers towards informed algorithm selection based on precision considerations.