no code implementations • 7 Mar 2024 • Lilian Ngweta, Mayank Agarwal, Subha Maity, Alex Gittens, Yuekai Sun, Mikhail Yurochkin
Large Language Models (LLMs) need to be aligned with human expectations to ensure their safety and utility in most applications.
no code implementations • 7 Dec 2023 • Felipe Maia Polo, Mikhail Yurochkin, Moulinath Banerjee, Subha Maity, Yuekai Sun
We develop methods for estimating Fr\'echet bounds on (possibly high-dimensional) distribution classes in which some variables are continuous-valued.
1 code implementation • 2 Oct 2023 • Subha Maity, Mayank Agarwal, Mikhail Yurochkin, Yuekai Sun
In this paper, we demonstrate that contrastive learning (CL), a popular variant of SSL, tends to collapse representations of minority groups with certain majority groups.
no code implementations • 13 Apr 2023 • Soham Bakshi, Subha Maity
In the special cases in which the Bayes decision rule is identified, we develop a simple algorithm to learn the Bayes decision rule, that does not require knowledge of the noise distribution.
1 code implementation • 20 Feb 2023 • Lilian Ngweta, Subha Maity, Alex Gittens, Yuekai Sun, Mikhail Yurochkin
Learning visual representations with interpretable features, i. e., disentangled representations, remains a challenging problem.
1 code implementation • 7 Jun 2022 • Subha Maity, Saptarshi Roy, Songkai Xue, Mikhail Yurochkin, Yuekai Sun
The benefits of overparameterization for the overall performance of modern machine learning (ML) models are well known.
1 code implementation • 26 May 2022 • Subha Maity, Mikhail Yurochkin, Moulinath Banerjee, Yuekai Sun
However, it is conceivable that the training data can be reweighted to be more representative of the new (target) task.
1 code implementation • 26 May 2022 • Subha Maity, Debarghya Mukherjee, Moulinath Banerjee, Yuekai Sun
Time-varying stochastic optimization problems frequently arise in machine learning practice (e. g. gradual domain shift, object tracking, strategic classification).
1 code implementation • ICLR 2021 • Subha Maity, Songkai Xue, Mikhail Yurochkin, Yuekai Sun
As we rely on machine learning (ML) models to make more consequential decisions, the issue of ML models perpetuating or even exacerbating undesirable historical biases (e. g., gender and racial biases) has come to the fore of the public's attention.
no code implementations • NeurIPS 2021 • Subha Maity, Debarghya Mukherjee, Mikhail Yurochkin, Yuekai Sun
Many instances of algorithmic bias are caused by subpopulation shifts.
no code implementations • 28 Sep 2020 • Subha Maity, Debarghya Mukherjee, Mikhail Yurochkin, Yuekai Sun
If the algorithmic biases in an ML model are due to sampling biases in the training data, then enforcing algorithmic fairness may improve the performance of the ML model on unbiased test data.
no code implementations • 23 Mar 2020 • Subha Maity, Yuekai Sun, Moulinath Banerjee
We study the minimax rates of the label shift problem in non-parametric classification.
1 code implementation • 26 Dec 2019 • Subha Maity, Yuekai Sun, Moulinath Banerjee
We consider the task of meta-analysis in high-dimensional settings in which the data sources are similar but non-identical.