Lecture
Optimizing Loss Functions: Gradient Descent Variants
Description
This lecture covers the concept of early stopping in gradient descent, emphasizing the importance of minimizing test error over loss function optimization. The instructor explains the use of subgradients for non-differentiable functions and introduces stochastic gradient descent as a faster alternative. Variants like momentum-based optimization and adaptive learning rates are discussed, highlighting the trade-offs between speed and convergence to global minima.
Official source
About this result
This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.