Explainable Machine Learning for IoMT Intrusion Detection: A Comparative Benchmark of PySpark and Scikit-learn

• A controlled comparison of matched PySpark and scikitlearn classifiers for multiclass IoMT intrusion detection.
• A joint assessment of predictive performance, training time, and prediction time under the same single-node conditions.
• Global and local explanations of the best-performing PySpark model using SHAP and LIME.