Repository for home work and project
My project for the class was done on General Adveserial Networks, in what I would call a gentle introduction to the cost function and some variations, e.g, Minimax, Non-Saturating, WGAN. The video for the presentation can be found here -> https://youtu.be/4BvQMSOvb9w.
Purpose: Intro to Python data structures. As mentioned in HW2 below, make sure you understand the basic data structures of python, e.g., tuples, lists, dictionaries, collections, etc.
Purpose: Implement your own version of GridSearch, really intended for those with less experience in Python coding and working with dictionaries within dictionaries, list comprehensions, etc. I have programming experience in other languages and didn't find python terribly difficult to pick up. Make sure you understand the basic data structures first!
Purpose: Decision Making With Matrices
Purpose: Train a convolutional neural network on MNSIT data and then apply transfer learning to a set of hand drawn letters (A through E). We compared letters drawn on lined paper vs cleaner images drawn directly on a touch screen laptop. As expected, the cleaner images had a higher validation accuracy.