Have questions on AWS? Join our office hours!
Our Research Enablement team will be hosting Zoom office hours to answer your questions and help you find out if AWS is right for you. We’ll be answering questions such as:
- Computations needs that are the best fit for AWS
- What are the costs
- How to request an account
- Live demo of how to use
- …and answer any of your questions!
What are others saying about DGIT’s AWS?
A scalable environment for AI model evaluation
The Hsu Lab, a part of the Medical & Imaging Informatics group in the Department of Radiological Sciences, is developing computational tools and methodologies for evaluating Artificial intelligence (AI)/machine learning (ML) models using simulated and real-world datasets. Their goal is to establish a platform that permits end users to test algorithms on a variety of scenarios, characterizing the bounds of model performance.
DGIT’s AWS provides a scalable platform for storing test datasets, running one or more AI/ML algorithms in parallel, and analyzing the evaluation runs. Using a large collection of images, Dr. Hsu’s team is able to evaluate the precision and recall of multiple AI/ML models to detect breast cancers in various subgroups. AWS provides flexibility for us to change instance types, launch additional instances, and leverage AWS managed services to optimize how evaluation experiments are set up and executed.