Department of Preventive Medicine Biostatistics Seminar Series: Harnessing Generative AI and Large Language Models for Revolutionary Statistical Practice

The Division of Biostatistics of the Department of Preventive Medicine, UTHSC, invites you to attend the following seminar.

Time: Monday, May 11, 2026, 2:00PM - 3:00PM CT

ZOOM Virtual Room Connection: Register in advance for this meeting to get the Zoom link

Seminar Website: https://www.uthsc.edu/preventive-medicine/events.php

Speaker Bio: https://www.rit.edu/directory/epfeqa-ernest-fokoue

Harnessing Generative AI and Large Language Models for Revolutionary Statistical Practice: Efficiency, Insight, and Human-AI Synergy

Ernest Fokoue, Ph.D.

Rochester Institute of Technology

The rapid evolution of Generative Artificial Intelligence and Large Language Models (LLMs) marks a paradigm shift in how statistical knowledge is accessed, applied, and communicated. Far beyond text completion tools, modern LLMs represent complex probabilistic systems rooted in statistical estimation, optimized through mechanisms such as tokenization, attention, and transformer architectures. These core components - where input sequences are discretized into tokens, and relationships between them are modeled through attention-weighted representations - enable LLMs to learn and generate coherent, context-aware language at scale. This presentation will seek to illuminate how statistical practitioners can meaningfully integrate these models into their workflows - not as replacements for human expertise, but as powerful allies for enhancing efficiency, deepening insight, and accelerating discovery. We begin by uncovering the statistical DNA of LLMs, tracing their lineage through conditional probability modeling, maximum likelihood estimation, and regularization. We then demonstrate practical use cases where LLMs can augment core statistical tasks - from exploratory data analysis and model interpretation to reproducible code generation and stakeholder-ready report writing. Live demonstrations will reveal how human-AI collaboration can supercharge clarity, reproducibility, and speed. I will also address ethical and methodological guardrails to ensure that LLMs are deployed responsibly within rigorous statistical frameworks. It is my hope and intention that attendees will leave equipped with conceptual clarity, practical templates, and a vision of the future where statistics and generative AI converge to form an unprecedented synthesis of intelligence - human and machine, rigorous and creative.