This session explores the core mathematical and algorithmic underpinnings of modern machine learning. Topics include supervised and unsupervised learning models, optimization strategies like gradient descent, and model evaluation metrics such as accuracy, precision, recall, and AUC. Emphasis will be placed on practical implementations using tools like scikit-learn and TensorFlow. The session will also examine current research trends in semi-supervised and self-supervised learning. Participants will gain insights into how foundational theories translate into real-world applications, with a focus on educational tools and open-source resources for ML learning and experimentation.