Supervised Learning Foundations | Regression, Classification & Leakage | SRAI

This lesson introduces the foundations of supervised machine learning through practical regression and classification examples. You will learn how to: • distinguish regression from classification; • prepare training, validation, and test datasets; • establish meaningful baseline models; • evaluate regression using RMSE; • evaluate classification using ROC AUC; • recognize and prevent data leakage; • compare predictive performance responsibly; • interpret model results in context. The accompanying demonstration uses reproducible Python workflows and a fixed random seed. It emphasizes disciplined validation, transparent comparison, and responsible interpretation—not merely obtaining an impressive metric. Key reference results presented in the lesson: • Regression test RMSE: 30.876 • Seasonal-naive RMSE: 40.496 • Training-mean RMSE: 103.958 • Classification ROC AUC: 0.955 • Leakage demonstration RMSE: 2.048 — rejected as invalid Production unit: PU-B02-C01 Book 2: Machine Learning Lesson 1: Supervised Learning Foundations GitHub repository: https://github.com/mbayekebe/SRAI_Boo... Author and instructor: Mbaye Kebe Statistical Research and Artificial Intelligence — SRAI #MachineLearning #SupervisedLearning #Regression #Classification #ResponsibleAI