Abstract by Neal Munson
Asynchronous Dynamic Neural Networks
Neural Networks, with Deep Learning in particular, has transformed the field of Artificial Intelligence, making unprecedented strides in image classification, image generation, and reinforcement learning. These state-of-the-art solutions are all synchronous and static by design. Comparatively, many real-world networks (specifically those in biology) are by nature asynchronous and dynamic. I will discuss some methods whereby asynchronous dynamic networks can be modeled, as well as some potential applications of utilizing dynamic asynchronous neural networks.