Fundamentals of neural networks

Presented a tech session at FAST-NUCES on the fundamentals of neural networks. The session was focused on building an intuitive understanding of why and how neural networks work & learn. We started with a very simple number multiplier network which when given a number, outputs the double of it, and ended with a multi-layer network capable of classifying images of numbers from MNIST data-set

The source code used in the session is available at this link

Neural Networks by Marib Sultan