Ian Goodfellow - Numerical Computation for Deep Learning - AI With The Best Oct 14-15, 2017
AI With The Best hosted 50+ speakers and hundreds of attendees from all over the world on a single platform on October 14-15, 2017. The platform held live talks, Insights/Questions pages, and bookings for 1-on-1s with speakers.
Ian Goodfellow is a Research Scientist at Google Brain and presented on Numerical Computation for Deep Learning.
This presentation covers chapter 4 of the Deep Learning textbook ( www.deeplearningbook.org ). Deep learning algorithms are usually described in terms of real numbers, with infinite precision. We usually implement these algorithms using 32-bit floating point numbers on GPU. While this may seem like an implementation detail, it can have a profound effect on the performance of the algorithm. Knowing how to write numerically precise code can be the difference between success and failure in deep learning. This presentation describes how to avoid the most common pitfalls.
http://withthebest.com/
/ withthebest
/ withthebestconf