Society of Statistical AI
The Department of Statistics is proud to lead the Society of Statistical AI at Purdue University, an initiative that brings together faculty dedicated to advancing artificial intelligence from a statistical perspective. Our aim is to harness and guide AI technologies in ways that are rigorous, data-driven, and impactful across science, engineering, health, and society. By working collaboratively with colleagues in computer science, engineering, and other disciplines, we position Purdue as a leader in shaping the future of AI, with Statistics playing a central role in ensuring that methods are reliable, interpretable, and grounded in data.
Faculty working with AI
Bruce Craig works on Statistical Consulting Service with AI. They have created their own private Knowledge Graph (KG)+Retrieval Augmented Generation (RAG) LLM platform that maintains the ethical and privacy concerns of their clients but hopefully enhances the efficiency of their service/consultants.
Shimeng Huang hsa focused research on developing innovative financial instruments using advanced statistical and machine learning methods to better manage the economic impacts of disasters.
Kiseop Lee is interested in AI-driven approaches to challenges in modern financial markets, including algorithmic trading, limit order book (LOB) analysis and prediction, and more. He is also exploring the application of prompt engineering in finance, particularly the use of large language models (LLMs) to support decision-making.
Michael Levine connects research with the world of AI lies through unsupervised learning and optimization.
Faming Liang has research interests that include uncertainty-aware AI, sparse deep learning, stochastic deep learning, reinforcement learning, and generative modeling.
Dennis KJ Lin has research that mainly focuses on "Data Quality"—in addition to the conventional study on Data Quality Assurance ("Data are Right"). He is more interested in how to collect "Right Data" (the data that will solve the target problems). That is highly relevant to Computer Experiment, Uncertainty Qualification, and Digital Twins.
Guang Lin has research interests that include: (1) Uncertainty Quantification and Interpretable, Robust Trustworthy LLM-based AI Models; (2) Interpretable, Reliable AI Models for Human Health & Disease Prediction.
Chuanhai Liu seeks the foundations of AI where problem-solving drives machines to recognize, create, and autonomously model the world.
Raghu Pasupathy’s research focus and expertise lie in aspects of optimization especially as they apply to AI/ML. This includes infinite-dimensional optimization, e.g., optimization over the space of probability measures, variational inference problems, and uncertainty quantification.
Vinayak A P Rao’s research interests include generative modeling, probabilistic inference and optimal decision-making, especially subject to constraints resulting from computational tractability, domain knowledge as well as ethical and societal considerations like privacy and fairness.
Walid K Sharabati's main interests are building language models for the digital assistants (data science behind virtual assistants), developing classification and classification algorithms (supervised and unsupervised learning).
Qifan Song has research interests including the mathematical foundations of the robustness and safety of AI models, trustworthiness of LLMs, computationally efficient algorithms and uncertainty quantification in modern data sciences.
Jianxi Su is an actuarial scientist by training, and he utilizes AI to solve problems arising in insurance and risk management contexts.
Wei Sun has research that focuses on reinforcement learning—particularly reinforcement learning with human feedback—and statistical considerations in large language model problems, including alignment, evaluation, agentic learning, and reasoning.
Lin Wang’s current specific interest is in the influence function and data valuation methods to diagnose and interpret AI models.
Xiao Wang has research that centers on machine learning and generative AI, nonparametric statistics, and functional data analysis with particular emphasis on developing methods for high-dimensional and complex data.
Mark Ward works on dozens of AI-focused research and development projects (from industry) in The Data Mine every year, with lots of real-world problems.
Bowei Xi has research that focuses on identifying the vulnerabilities of AI models and improving AI safety from a statistical perspective.
Jun Xie has research that focuses on causal AI, exploring how causal inference, machine learning, and AI methods can enhance each other. She also develops causal generative AI methods for treatment effectiveness evaluation and precision medicine.
Fei Xue has research interests including data integration, statistical inference, and mobile health data.
Lingsong Zhang is specifically interested in generated AI and causal inference. In addition, he is interested in sports-related statistics work that needs AI tools as well.