Spring Seminar Series – Dr. Irene Chen

Beyond Bias Audits: Bringing Equity to the Machine Learning Pipeline

Dr. Irene Chen

Microsoft Research

3/31/2023

11AM – 12PM CST

WebEx: https://utsa.webex.com/utsa/j.php?MTID=m97203d6036ae3b1f22fd0fb98da3c2d1

Advances in machine learning and the explosion of clinical data have demonstrated immense potential to fundamentally improve clinical care and deepen our understanding of human health. However, algorithms for medical interventions and scientific discovery in heterogeneous patient populations are particularly challenged by the complexities of healthcare data. Not only are clinical data noisy, missing, and irregularly sampled, but questions of equity and fairness also raise grave concerns and create additional computational challenges.

In this talk, I present two approaches for leveraging machine learning towards equitable healthcare. First, I examine how to incorporate differences in access to care into the modeling step. Using a deep generative model, we examine the task of disease phenotyping in heart failure and Parkinson’s disease. Second, I demonstrate how to address one specific health disparity through the early detection of intimate partner violence from clinical indicators. Using a time-based model with noisy labels, we can correct for biases in data measurement to learn more clinically useful subtypes and improve prediction. The talk concludes with a discussion about how to rethink the entire machine learning pipeline with an ethical lens to building algorithms that serve the entire patient population

Special Seminar Series – Dr. Nancy Chen

Neural Language Generation: From Self-Supervised Representation Learning to Multimodal Contextualization

Dr. Nancy Chen

Infocomm Research (I2R), A*STAR (Agency for Science, Technology, and Research), Singapore

Wednesday, April 5, 2023

9AM – 10AM CST

WebEx: https://utsa.webex.com/utsa/j.php?MTID=m66139b13650b563509afee98c46142e9

Abstract:

The recent rapid advancements of neural language models have awed both domain experts and laymen alike. Such language models can fuel many natural language generation applications, including summarization, conversational agents, and machine translation. In this talk, we will illustrate how to harness the capabilities of these language models from a research perspective to address on-going technical challenges. In particular, we will delve into four research themes: scalable data augmentation, factually consistent generation, controllable neural modelling, and multimodal contextualization.

Spring Seminar Series – Dr. Hongseok Namkoong

A New Computation-driven Framework for Adaptive Experimentation

Hongseok Namkoong, PhD

Assistant Professor in the Decision, Risk, and Operations division

Columbia Business School

4/21/2023

11 am – 12pm CST

WebEx: https://utsa.webex.com/utsa/j.php?MTID=m568fe3d73989cb26d676f896e952c3c0

Abstract:  Experimentation serves as the foundation of scientific decision-making. Adaptive allocation of measurement effort can significantly improve statistical power. However, implementing standard bandit algorithms, which assume continual reallocation of measurement effort, is challenging due to delayed feedback and infrastructural or organizational difficulties. To address this, we introduce a new framework for adaptive experimentation, motivated by practical instances involving a limited number of reallocation epochs in which outcomes are measured in batches. Our framework diverges from the traditional theory-driven paradigm by utilizing computational tools for algorithmic design.

We observe that normal approximations, which are universal in statistical inference, can also guide the design of scalable adaptive designs. By deriving an asymptotic sequential experiment, we formulate a dynamic program that can leverage prior information on average rewards. We propose a simple iterative planning method called Residual Horizon Optimization, which selects sampling allocations by optimizing a planning objective. Our method significantly improves statistical power over standard adaptive policies, even when compared to Bayesian bandit algorithms (e.g., Thompson sampling) that require full distributional knowledge of individual rewards. Overall, we expand the scope of adaptive experimentation to settings that pose difficulties for standard adaptive policies, including problems with a small number of reallocation epochs, low signal-to-noise ratio, and unknown reward distributions.

This work was led by Ethan Che. Paper link: https://arxiv.org/abs/2303.11582.