11 papers
Interpretable Generalized Additive Models for Datasets with Missing Values
Hayden McTavish, Jon Donnelly, Margo Seltzer +1
Many important datasets contain samples that are missing one or more feature values. Maintaining the interpretability of machine learning models in the presence of such missing dat…
It's LIT! Reliability-Optimized LLMs with Inspectable Tools
Ruixin Zhang, Jon Donnelly, Zhicheng Guo +4
Large language models (LLMs) have exhibited remarkable capabilities across various domains. The ability to call external tools further expands their capability to handle real-world…
This EEG Looks Like These EEGs: Interpretable Interictal Epileptiform Discharge Detection With ProtoEEG-kNN
Dennis Tang, Jon Donnelly, Alina Jade Barnett +8
The presence of interictal epileptiform discharges (IEDs) in electroencephalogram (EEG) recordings is a critical biomarker of epilepsy. Even trained neurologists find detecting IED…
Doctor Rashomon and the UNIVERSE of Madness: Variable Importance with Unobserved Confounding and the Rashomon Effect
Jon Donnelly, Srikar Katta, Emanuele Borgonovo +1
Variable importance (VI) methods are often used for hypothesis generation, feature selection, and scientific validation. In the standard VI pipeline, an analyst estimates VI for a…
Leveraging Predictive Equivalence in Decision Trees
Hayden McTavish, Zachery Boner, Jon Donnelly +2
Decision trees are widely used for interpretable machine learning due to their clearly structured reasoning process. However, this structure belies a challenge we refer to as predi…
Rashomon Sets for Prototypical-Part Networks: Editing Interpretable Models in Real-Time
Jon Donnelly, Zhicheng Guo, Alina Jade Barnett +3
Interpretability is critical for machine learning models in high-stakes settings because it allows users to verify the model's reasoning. In computer vision, prototypical part mode…