Mike Hughes - Machine Learning Research

Welcome! My research spans statistical machine learning and its applications in healthcare and the sciences.

I am an Assistant Professor in the Dept. of Computer Science at Tufts University and a primary faculty member for the Tufts Machine Learning research group. Previously, I was a postdoctoral fellow in computer science at Harvard SEAS, advised by Prof. Finale Doshi-Velez. I did my Ph.D. in CS at Brown University in May 2016, advised by Prof. Erik Sudderth (now at UC-Irvine).

Research Highlights

My lab pursues advances to core machine learning methods as well as high-impact applications.

First, I'm interested in fundamental problems in statistical machine learning:

  • probabilistic models for multivariate time-series with irregular or missing data
  • scalable inference for latent variable models (see our AABI 2019 paper)
  • effective semi-supervised learning with latent variable models (see our AISTATS 2018 paper)
  • Bayesian nonparametric approaches to growing model complexity as data demands (see our BNPy Python package and NeurIPS 2015 paper)
  • model-based reinforcement learning with open-world, partially-observable state (see our AISTATS 2020 paper)
  • optimizing deep learning models to be more interpretable (see our AAAI 2020 and AAAI 2018 papers)

Second, I'm interested in several exciting clinical applications of these techniques:

  • antidepressant drug recommendations for patients with major depression
  • forecasting deterioration and need for interventions in the Intensive Care Unit (ICU)
  • forecasting chemotherapy risks for acute myeloid leukemia patients

For more, see my Research page