Deep Learning from Scratch using Python
Details
Abstract:
Many of us have used libraries like Keras and TensorFlow to train Deep Neural Networks, but few of us fully understand what is going on under the hood.
In this talk, we'll walk through why Deep Learning is so effective at solving image classification tasks from scratch using Python. We'll start by implementing the layers of our network using classes. We will fully define the backpropogation algorithm such that the layers work correctly together, before working through a number of optimization techniques such as Dropout, Momentum, and Weight Regularization. We'll end with a Jupyter notebook running a flexible deep learning framework live that can then be extended to create arbitrarily deep networks.
Bio:
Seth Weidman loves helping organizations best leverage data assets - and teaching others how to do the same.
Seth joined Metis from Trunk Club, where he built their lead scoring models and recommender systems. He majored in Math and Econ at The University of Chicago, and outside of work he likes to play Ultimate Frisbee while also making time to enjoy local art, comedy, and theater scenes.
Schedule:
6:00 - 6:30 Food and Drinks
6:30 - 7:30 Session
7:30 - 8 Shmooze
