Introduction to Artificial Neural Networks and Gradient Descent

Teaching material for an advanced 2h course on neural networks I gave at the International Interdisciplinary Computational Cognitive Science Spring School in Freiburg, 2019 (http://iiccsss.org/)

Contents

The course introduces artificial neural networks as composition of linear and nonlinear functions and is divided into three sections:

1. Maths Refresher
2. Basics
   1. Linear Regression
   2. Logistic Regression
   3. (Stochastic) Gradient Descent
3. Neural Networks
   1. Architecture
   2. Backpropagation Algorithm

I provide code examples in Python for the second and third section

Material

• code
  contains code examples implemented as iPython notebooks

• slides
  lecture slides

Contact

If you spot typos, have suggestions or would like to use the material, send me an email (firstname (dot) lastname (at) psy (dot) ox (dot) ac (dot) uk)