Radically improving neural networks with insights from modern neuroscience

Please join PyData Pittsburgh for the talk Radically improving neural networks with insights from modern neuroscience by Dr. Rorry Brenner, founder and CEO of Perforated AI! We'll be gathering in person at the Swartz Center for Entrepreneurship at Carnegie Mellon University. Food and beverages will be provided.

About the talk

The architecture and mechanics of the current generation of deep neural networks are based on our understanding of biological neurons from the 1950s. Neuroscience has advanced a lot since then, but our approach to building and training artificial neural networks hasn't kept up! Dr. Rorry Brenner has developed and patented a new algorithm that brings insights from modern neuroscience to silicon-based deep learning. Dr. Brenner's algorithm, called perforated backpropagation, has been shown to deliver significant accuracy and efficiency improvements to existing neural network models across a wide variety of fields and applications.

In this talk, Dr. Brenner will:

• Review the biological inspiration behind the original artificial neuron
• Survey more recent developments in neuroscience that explain how biological neurons can exponentially outperform their silicon-based counterparts
• Provide an overview of how the technology he created brings this more sophisticated understanding of neurons to machine learning

Finally, Dr. Brenner will share some examples of applications that demonstrate the benefits of his approach and discuss potential opportunities to try out the technology as a beta tester.

Getting to the event

The Swartz Center for Entrepreneurship is located within the Tepper School of Business on the Carnegie Mellon University campus. After you arrive at the Tepper Quad, enter the main doors of the Tepper School building (on floor 2). The Swartz Center for Entrepreneurship is located on the third level (on floor 3) in Suite 3700.

Parking is available in the East Campus Garage on Forbes and Beeler Streets.

For more information about directions and parking, please see here.

Learn more about PyData Pittsburgh and subscribe to announcements and event updates at https://news.pypgh.org.