Data + AI
Feature-Relevant Data Reduction
A research project on lossy compression for scientific computing, focused on preserving the features that actually matter.
This work explored whether large simulations could shrink dramatically without losing the signal researchers care about, and was presented at SuperComputing 2018.