Zachery Boner

I am a fourth-year PhD Candidate at Duke University, advised by Dr. Cynthia Rudin, researching the science of simple machine learning models. I believe that the inability to reliably identify the patterns and biases encoded by black box machine learning models creates the potential for harm in many high-stakes prediction tasks. My primary research objective is to characterize when and why simple models are competitive with black box models in as general a setting as possible. My goal is to rigorously upend the frequently-false belief that accuracy and interpretability are in opposition for many high-stakes decision domains, and to apply this intuition to develop robust and interpretable machine learning models for medical and other high-stakes applications. I am fortunate to be funded by the NSF Graduate Research Fellowship.