Madalina (Ina) Fiterau’s research lies at the intersection of machine learning and healthcare. She joined the UMass Amherst College of Information and Computer Sciences in 2018, after completing a joint computer science and bioengineering postdoctoral appointment at Stanford University. Fiterau earned her PhD in machine learning from Carnegie Mellon University.
Her research focuses on devising methods that are suitable for handling multi-modal, heterogeneous data sets -- time series, videos, images, and more -- to construct models that can handle various types of data jointly. Fiterau’s Information Fusion Lab is currently working on a project combining features extracted from brain MRIs with patient demographics, test results, and contextual information, to detect Alzheimer’s disease earlier than traditional diagnostics can.
Fiterau’s research is applicable in many fields beyond healthcare. Her models can handle any scenario where data is collected in multiple modalities. “It’s about getting more out of your existing data by leveraging the often-overlooked components, as opposed to collecting much more training data,” she says.
“I’ve always liked math and programming, so I like the intellectual challenges. I start with a concrete problem, abstract it, work at a higher-level space, and come up with a solution that appears to magically work. Except it’s not magic, it’s just math.”
Follow Fiterau on Twitter at @mfiterau.