Presented By O’Reilly and Intel Nervana
Put AI to work
September 17-18, 2017: Training
September 18-20, 2017: Tutorials & Conference
San Francisco, CA

All the linear algebra you need for AI

Rachel Thomas (fast.ai)
2:35pm–3:15pm Wednesday, September 20, 2017
Implementing AI
Location: Imperial A Level: Intermediate
Secondary topics:  Deep learning, Tools and frameworks
Average rating: **...
(2.00, 6 ratings)

Prerequisite Knowledge

  • A working knowledge of Python and the Jupyter Notebook

What you'll learn

  • Learn how to work with tensors
  • Understand some of the key techniques for implementing AI, including the chain rule, broadcasting, and principal component analysis
  • Gain exposure to PyTorch

Description

If the math used in AI seems intimidating, this tutorial is for you. Rachel Thomas walks you through working with arrays of different dimensions and how broadcasting handles data dimensions. As a fun application of linear algebra, you’ll learn how to create a deep neural network from scratch and use it to recognize handwritten digits. You’ll also gain hands-on experience with PyTorch, the Python framework for GPU computing developed by Facebook.

Photo of Rachel Thomas

Rachel Thomas

fast.ai

Rachel Thomas is the cofounder of fast.ai and a researcher in residence at USF Data Institute, where she teaches numerical linear algebra. Rachel helped create the free Practical Deep Learning for Coders MOOC, which 50,000 students have started. Previously, she worked as a quant in energy trading, a data scientist and engineer at Uber, and a senior instructor at Hackbright. Rachel is a popular writer on data science and diversity in tech. Her writing has made the front page of Hacker News and Medium, has been included in newsletters by O’Reilly, Fortune, crunchbase, and Mattermark, and has been translated into Spanish, Portuguese, and Chinese. Rachel holds a PhD in mathematics from Duke.