Deep Learning from First Principles

A from-scratch-as-possible machine learning course with a strong focus on deep learning theory and techniques.

Artificial Neural Networks

• Introduction and the perceptron
• Feedforward Neural Networks
• Backpropagation and gradient descent
• Cross Entropy and He Initialization

Convolutional Neural Networks

• Convolution
• Convolution and pooling with im2col
• Convolution Backpropagation
• LeNet-style CNN
• CNN Improvements

License

© 2025 Andrea Giardina. This work is licensed under CC BY 4.0.