Výsledky vyhledávání - "Nunnally, Mark"

Report

LLMs Understand Glass-Box Models, Discover Surprises, and Suggest Repairs

Autor: Lengerich, Benjamin J., Bordt, Sebastian, Nori, Harsha, Nunnally, Mark E., Aphinyanaphongs, Yin, Kellis, Manolis, Caruana, Rich

We show that large language models (LLMs) are remarkably good at working with interpretable models that decompose complex outcomes into univariate graph-represented components. By adopting a hierarchical approach to reasoning, LLMs can provide compre

Externí odkaz: http://arxiv.org/abs/2308.01157

Zobrazit plný text záznamu

Report

Estimating Discontinuous Time-Varying Risk Factors and Treatment Benefits for COVID-19 with Interpretable ML

Autor: Lengerich, Benjamin, Nunnally, Mark E., Aphinyanaphongs, Yin, Caruana, Rich

Treatment protocols, disease understanding, and viral characteristics changed over the course of the COVID-19 pandemic; as a result, the risks associated with patient comorbidities and biomarkers also changed. We add to the conversation regarding inf

Externí odkaz: http://arxiv.org/abs/2211.08991

Zobrazit plný text záznamu

Report

Interpretability, Then What? Editing Machine Learning Models to Reflect Human Knowledge and Values

Autor: Wang, Zijie J., Kale, Alex, Nori, Harsha, Stella, Peter, Nunnally, Mark E., Chau, Duen Horng, Vorvoreanu, Mihaela, Vaughan, Jennifer Wortman, Caruana, Rich

Machine learning (ML) interpretability techniques can reveal undesirable patterns in data that models exploit to make predictions--potentially causing harms once deployed. However, how to take action to address these patterns is not always clear. In

Externí odkaz: http://arxiv.org/abs/2206.15465

Zobrazit plný text záznamu

Report

GAM Changer: Editing Generalized Additive Models with Interactive Visualization

Autor: Wang, Zijie J., Kale, Alex, Nori, Harsha, Stella, Peter, Nunnally, Mark, Chau, Duen Horng, Vorvoreanu, Mihaela, Vaughan, Jennifer Wortman, Caruana, Rich

Recent strides in interpretable machine learning (ML) research reveal that models exploit undesirable patterns in the data to make predictions, which potentially causes harms in deployment. However, it is unclear how we can fix these models. We prese

Externí odkaz: http://arxiv.org/abs/2112.03245

Zobrazit plný text záznamu