Implementing a Neural Network from Scratch

Rodrigo Girão Serrão

Data Science Deep Learning Engineering Machine-Learning Science

See in schedule: Tue, Jul 27, 09:30-11:00 CEST (90 min)

The fact that you can just install a library and train machine learning models straight away is amazing, to say the least.
However, if your purpose is to learn how the models work, writing `import pytorch` will not do much for you.

In this hands-on workshop we implement a basic neural network framework, dealing with all the nitty-gritty details, and making sure you build something amazing in a surprisingly short amount of code. We also apply our framework to a real classification problem.

Attendees are expected to have working knowledge of neural networks, matrix algebra and differentiation.
The only installation requirements are Python 3 and numpy.

Type: Training (180 mins); Python level: Intermediate; Domain level: Beginner


Rodrigo Girão Serrão

Rodrigo has always been fascinated by problem solving and that is why he picked up programming – so that he could solve more problems.
He also loves sharing knowledge, and that is why he spends so much time writing articles @ https://mathspp.com, giving workshops and courses, and writing books.

His main areas of scientific interest are mathematics (numerical analysis in particular) and programming in general (with a preference for the Python and APL languages), but Rodrigo also enjoys reading fantasy books, watching silly comedy movies and eating chocolate.