After over ten years of research primarily in robotics, Ebert-Uphoff's research now focuses primarily on Bayesian Networks, a data mining tool at the intersection of Artificial Intelligence and Statistics. Bayesian networks have gained much popularity in recent years as a tool for modeling systems that contain uncertainty. They have been applied to a great range of systems, including systems in nature (for example, sea breeze and volcano eruption models), mechanical systems (for example, failure analysis for printers), medical systems (such as cancer diagnosis) and information systems (e-mail spam filters).

Ebert-Uphoff's current interests include the calculation and visualization of link strengths and connection strengths in Bayesian networks. She is also interested in causal discovery techniques using Bayesian networks and application of those techniques to derive knowledge of causal relationships in real-world systems based on data. Anyone interested in trying some of those techniques for their own system is encouraged to contact her for potential collaboration.