Watch Sean Sall, Machine Learning Scientist at Arterys, present “Intro to Neural Networks with NumPy, Theano, and Keras”.
You can follow along with Sall and use his Github repository to build a shallow neural network that can learn arbitrary mathematical functions.
As Sall explains:
“This repository is intended to present a highly accessible introduction to neural networks, building upon first principles. This introduction will build up from simple linear regression with the end goal of providing a reasonably comprehensive picture of how, why, and when neural networks work.”
The Kaizen Data Conference is an applied data conference focused on data analytics, data processing, data management, data visualization and machine learning. Hosted at Galvanize’s San Francisco campus, the event brought together data science thought-leaders from top companies in the Bay area, including Slack, Instacart, Uber, Pinterest, Netflix, StitchFix, and more.