Learn the Math Behind Neural Networks!

DATE: 27 March 2023

The course is designed and delivered by Prof Claire David (York University, ATLAS Experiment) https://www.yorku.ca/science/clairedavid/

DESCRIPTION
Artificial Intelligence and Machine Learning is extremely popular and growing in many areas: academia, industry, not to mention our day-to-day life. It is thus important for scientists of all levels to get a basics understanding of the main algorithms. This day-workshop will explain to you step by step the mathematical operations behind a simple neural network (two morning sessions). In the afternoon session, we will code together a small neural network in python… by hand! That is to say, without any dedicated Machine Learning libraries! Don’t worry: it will be guided. After this course, you will have the foundations to explore more complicated neural network architectures and above all, you will not say “It’s Magical” but rather “It’s Mathematical!” PREREQUISITES From undergrads to emeritus professors, everyone is welcome. Bring a pen, some paper as you will actually do some of the calculations yourself (not much, promised) and a sense of humour.

NOTE
If you already know about neural networks but are curious about a basic implementation in python, you can only join the afternoon session.

If you are interested in the afternoon session but cannot make it, I can give you the pythonic neural network exercise and later share the solutions with you per email upon request.

SESSION 1
09:00 – 10:30 Theory behind Neural Networks Tutorial Rooms (6th Floor), UJ on Empire. Corner Empire & Napier Roads, Richmond, Johannesburg.

11:00 – 12:30 Theory behind Neural Networks Tutorial Rooms (6th Floor), UJ on Empire. Corner Empire & Napier Roads, Richmond, Johannesburg.

SESSION 2
14:00 – 16:00 Programming Tutorial Rooms (6th Floor), UJ on Empire. Corner Empire & Napier Roads, Richmond, Johannesburg.

16:00 – 17:30 DRINKS & SNACKS 8th Floor, UJ on Empire Corner Empire & Napier Roads, Richmond, Johannesburg.

Please RSVP before 23 March 2023 to Prof Charis Harley (charley@uj.ac.za). You may choose which session you would like to attend, so please specify whether you will join us for Session 1, Session 2, or both. We will only be able to accommodate 20 individuals – acceptance will be on a first come, first serve basis. Preference will be given to those attending both sessions.