Comparative Study of Machine Learning Algorithms for E-mail Spam Detection

Email spam is still a continuous issue that affects user experience, costs resources, and makes fraud and phishing possible. In addition to proposing an experimental pipeline and offering a repeatable methodology for model training, evaluation, and comparison, this work examines both traditional and contemporary machine learning approaches to email spam detection. We implement and examine a number of algorithms, including Multinomial Naive Bayes, Logistic Regression, Support Vector Machines, Decision Trees, Random Forests, a basic deep-learning baseline (bi-LSTM), using popular datasets (Enron, Spam Assassin, Ling-Spam) and standard preprocessing (cryptography, TF–IDF, header-feature extraction). We explore the interactions between performance, interpretability, and computing cost and give evaluation measures (accuracy, precision, recall, F1-score, ROC-AUC). Deployment issues, constraints, and future research goals are discussed in the paper's conclusion.