Skip to content
This repository was archived by the owner on Feb 27, 2021. It is now read-only.

epochml/epoch-neural-nets-intersession

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

65 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Introduction to Neural Networks

This repository contains the code and Colab notebooks for the Epoch "Introduction to Neural Networks" intersession, led by Ethan Haque and Dev Singh.

While these notebooks may run on environments outside of Colab, they do use some specific Colab libraries. Therefore, it is reccomended to use Colab to run these notebooks. Any notebooks that have architectures which greatly benefit from GPU acceleration have GPU acceleration enabled in the notebooks.

Note about Calculus

These notebooks assume that someone has no Calculus experience. While this means that we explain some things like backprop, gradient descent, etc. in easier-to-understand ways, it also means that we can't go as deep into ML math (like "Why is this activation function not a good idea for this purpose" or "What is the vanishing gradient problem?"). The curriculum is written assuming knowledge of MI4, so we can't explain those. For those who have a working knowledge of calculus, i.e. have taken at least BC1, preferably BC2 as well, and some intuition of how MVC might work, you may benefit from additional instruction in some ML/CS theory. The ML Cheatsheet is a good resource for those interested.

Schedule

Day 1 - Introduction

  • What is a Neural Network?
    • NN vs. ML vs. AI vs. Deep Learning
    • Goes over some of the mathematical concepts behind neural networks
    • Explain in depth how linear/dense neurons work
    • Implement an MNIST-classifying NN using just numpy

Day 2 - Implementing real-world neural networks

Day 3 - A deeper dive

About

Epoch introduction to neural networks intersession, SY2021

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •