Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore the fundamental algorithm driving machine learning in this 40-minute video lecture. Delve into the concept of backpropagation, deriving it from first principles. Begin with a historical background before tackling the curve fitting problem. Compare random and guided adjustments, then progress to derivatives and gradient descent. Examine higher dimensions and gain intuition on the chain rule. Investigate computational graphs and automatic differentiation. Conclude with a comprehensive summary and access additional resources, including Andrej Karpathy's playlist and Jürgen Schmidhuber's blog on backpropagation history. Enhance your understanding of this crucial machine learning concept through clear explanations and practical examples.
Syllabus
Introduction
Historical background
Curve Fitting problem
Random vs guided adjustments
Derivatives
Gradient Descent
Higher dimensions
Chain Rule Intuition
Computational Graph and Autodiff
Summary
Shortform
Outro
Taught by
Artem Kirsanov