Tin Nguyen is a PhD Student at MIT. He is advised by Tamara Broderick. His research develops computational methods for inference and sensitivity detection in Bayesian inference.
Publications
(2022).
Are you using test log-likelihood correctly?.
In ICBINB Workshop at NeurIPS 2022.
(2022).
Many processors, little time: MCMC for partitions via optimal transport couplings.
In AISTATS 2022.
(2022).
Measuring the robustness of Gaussian processes to kernel choice.
In AISTATS 2022.
(2020).
Approximate cross-validation for structured models.
In NeurIPS 2020.
Skills
R
3 years experience
Python
4 years experience
SQL
1 year experience
Experience
Research Intern
IBM Research
Reduced runtime of approximate cross-validation for structured models including hidden Markov models and conditional random fields.
Research Intern
IBM Research
Developed specification tests to reject the null hypothesis that neural networks trained on clean image data are well-specified for corrupted image data.
Awards
Student Poster Award
Outstanding Reviewer Award
Student Travel Award