Using Nengo and the Neural Engineering Framework to Represent Time and Space

Schedule
Wednesday, July 15, 2020
16:00 UTC
Terry Stewart

PhD

University of Waterloo and National Research Center Canada

Host: BCCN Berlin lectures series

Event Information

Host

BCCN Berlin lectures series

Duration

70 minutes

Abstract

The Neural Engineering Framework (and the associated software tool Nengo) provide a general method for converting algorithms into neural networks with an adjustable level of biological plausibility. I will give an introduction to this approach, and then focus on recent developments that have shown new insights into how brains represent time and space. This will start with the underlying mathematical formulation of ideal methods for representing continuous time and continuous space, then show how implementing these in neural networks can improve Machine Learning tasks, and finally show how the resulting systems compare to temporal and spatial representations in biological brains.

We use essential cookies to run the site. Analytics cookies are optional and help us improve World Wide. Learn more.