Dr Niemier's Lab

We are interested in studying how emerging technologies and architectures can ultimately impact machine learning (ML) workloads, like recommendation systems. While a recommendation system is a rather large problem to study, in the context of new technologies and/or architectures, models such as few-shot learning with memory augmented neural networks (MANNs) can be a good “proxy” for said problems, as many of the technology-enabled solutions for MANNs can also be used in recommendation systems.

Our overall goal is to study the strengths and weaknesses of certain systems based on the materials and architectures used to support them.