Brown Professor of Physics Stephon Alexander and other colleagues at Brown have been awarded a 2021 Seed Research Award for a project entitled, “Finding the Physics that Matters in Astrophysical and Astro-Particle Analyses with Interpretable Machine Learning.”
The group proposes to use interpretable machine learning to create a software framework for identifying and understanding the biases induced in the models scientists utilize by the training data they use. Scientists are often presented with a tradeoff between abundant but low-resolution data, and scarce high-resolution data. The aim of the project is to correct the biases in the abundant low-resolution data so that it is more like the scarce high-resolution data in the ways that affect the predictions of the machine learning models trained on that data. If successful, scientists would have access to higher quality, abundant data.
Assistant Professor of Computer Science Stephen Bach, Professor of Physics Ian Dell’Antonio, Hazard Professor of Physics Richard Gaitskell, and Assistant Professor of Physics Jonathan Pober will serve as Co-PI’s on the project.