ML Foundations III - Deep Learning

Learn the fundamental concepts of deep learning through this comprehensive lecture covering artificial neurons, activation functions, and feedforward neural networks. Master the mathematical foundations including calculus review of the chain rule and backpropagation algorithms for training deep neural networks. Explore key training algorithm considerations and discover the motivation behind Convolutional Neural Networks (CNNs). Understand kernels and convolution operations, examine CNN structure and pooling techniques, and analyze example CNN architectures with their practical use cases.

Syllabus

Artificial Neurons
Activation Functions used in Deep Neural Networks
Feedforward Neural Networks
Calculus Review: the Chain Rule
Backpropagation: Training Deep Neural Networks
Deep Learning Training Algorithm Considerations
Motivation for CNNs
Kernels and the Convolution Operation
Convolutional Neural Network Structure
Pooling
Example CNN Architectures and Use Cases