Machine Learning Essentials

Build a solid mathematical foundation for neural networks by learning the core concepts behind Machine Learning and how they connect.

What you’ll learn

  • Implement and interpret linear and logistic regression from scratch.
  • Use maximum likelihood estimation (MLE) to understand and guide learning.
  • Decide between supervised and unsupervised methods based on data and goals.
  • See how MLE maps directly to losses: MSE for linear, cross-entropy for logistic.

Hands-on application

  • Work with real datasets: clean, split, and structure them correctly.
  • Select appropriate objective functions for regression vs. classification.
  • Train with gradient descent and evaluate using RMSE, accuracy, precision/recall, ROC-AUC.

Prerequisites

  • Basic calculus (derivatives) and intro statistics (probability distributions).
  • Some programming experience (Python preferred).

Who it’s for

  • Developers new to ML who want a solid mathematical base before deep learning.
  • Builders who learn best by starting from first principles.

Format

  • Duration: ~4–6 hours, fully self-paced.
  • Structure: 6 bite-sized modules → short lesson, worked example, then a notebook to implement.