About Me

I am a researcher and graduate student in the DREAM Lab at UMass Amherst working with Gerome Miklau and Dan Sheldon. Before this, I was a Data Scientist at Allstate, an affiliate researcher at the Fiscal Research Center at Georgia State University, and studied mathematics and logic at Miami University in Oxford, Ohio.

Click here to view my CV.

You may contact me at brettcmullins(at)gmail.

This radio program on Frank Ramsey influenced my thinking and interests greatly.

Research Interests

Differential Privacy, Explainable Machine Learning, Learning Theory,
Causation, Formal Epistemology

My research involves using tools from mathematics, logic, and machine learning to approach a problem from a novel perspective.

News

Feb 2025 I’m teaching Intro to Programming at UMass Amherst this semester
Dec 2024 Our paper Efficient and Private Marginal Reconstruction with Local Non-negativity to appear at NeurIPS 2024
Nov 2024 I’m giving a talk at Miami University on Efficient Marginal Reconstruction under Differential Privacy
Aug 2024 Two of our papers to appear at TPDP 2024 - Efficient and Private Marginal Reconstruction with Local Non-negativity and Joint Selection Adaptively Incorporating Public Information for Private Synthetic Data
Apr 2024 Our paper Joint Selection Adaptively Incorporating Public Information for Private Synthetic Data appeared at AISTATS 2024
Oct 2023 Gave a talk on error bounds for private query answering at the UMass CS Theory Seminar
Oct 2023 Gave a talk introducing differential privacy and synthetic data at the UMass Learning Learning Seminar
Sept 2023 Our paper Quantifying Uncertainty of Unsupported Linear Queries for Private Query Release appeared at TPDP 2023