Neural networks through the lens of evolutionary dynamics
Dan C. Baciu (TU Delft - Architecture and the Built Environment, UCSB, ASB)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
This article revisits Artificial Neural Networks (NNs) through the lens of Evolutionary Dynamics. The two most important features of NNs are shown to reflect the two most general processes of Evolutionary Dynamics. This overlap may serve as a new and powerful connection between NNs and Evolutionary Dynamics, which encompasses a body of knowledge that has been built over multiple centuries and has been expanded to inspire applications across a vast range of disciplines. Consequently, NNs should also be applicable across the same range of disciplines—that is, much more broadly than initially envisioned. The article concludes by opening questions about NN dynamics, based on the new connection to Evolutionary Dynamics.