Research in the IDEA Lab focuses on developing and applying data science, artificial intelligence, and machine learning techniques with a focus on high-impact real-world applications. Our research foci include the development of:
- Autonomous data science techniques that can learn from real-world data
- Applied data science/artificial intelligence/machine learning techniques that can be deployed in real-world settings
- Intelligent autonomous agents that can successfully interact and learn while embedded in the real world
Success in dynamic, spatiotemporally varying, multi-agent environments requires the following four key components.
- Interaction: Data scientists must interact closely with domain scientists to ensure that the techniques are doing what is needed. The results of the methods must be communicated clearly and efficiently. Embedded learning agents must also be able to interact with the environment by taking actions and observing the results of its action. The environment should provide feedback, either implicitly or explicitly.
- Discovery: Data science methods and embedded agents both need to discover the most salient patterns from the overwhelming amount of information available. Success in complex environments requires an agent to focus its attention on the essentials. This focus can help domain scientists succeed by discovering new knowledge about complex phenomena.
- Exploration: An agent must balance the need to exploit its current knowledge about the task at hand with a need to actively explore alternative solutions. Curiosity is a required component of any successful agent.
- Adaptation: Embedded learning agents must be able to adapt to changes in the environment as well as to changes within themselves (such as hardware degradation). Applied data science methods must be adapted to best fit the requirements of the domain scientists.
General areas of research include:
- Autonomous pattern discovery: We develop methods to autonomously discover the most salient patterns or testable hypotheses in complex data sets. We have a special interest in large spatiotemporally varying numeric data.
- Knowledge representation: Real-world data is complex, large, and dynamic. We create knowledge representations for spatially and temporally varying relational data where agents will be able to reason about objects and relationships and the evolution of their environment over time and space.
- Robust machine learning techniques: We combine many approaches to machine learning to create robust learning algorithms. This includes techniques drawn from reinforcement learning, supervised learning, and relational learning.
Our real-world application areas provide an opportunity to make a significant difference outside of academia. Current application areas follow.
- High-impact weather prediction/weather analytics: In collaboration with researchers at the University of Oklahoma's School of Meteorology, the National Severe Storms Laboratory, and the Cooperative Institute for Mesoscale Meteorological Studies, we are developing novel weather analytics techniques for high-impact weather data. These techniques can improve the prediction and understanding of severe weather events, including tornados, thunderstorms, wind, and hail.
- Robotics: In collaboration with robotics researchers, we are developing novel approaches to autonomously creating robust task-oriented knowledge.
- STEM education: We are actively working to improve the teaching of Computer Science and Computational Thinking at all levels of education, including outreach to K-12 through robotics and programming and through the development of Computational Thinking classes for teachers and students in Oklahoma.
- Improving diversity in CS: McGovern has worked to improve the diversity of CS since she became a CS major. The lab participates actively in outreach events aimed at improving diversity including K-12 events, REU mentoring, and in being a welcoming space to all students.
- Space: We have a long-term interest in applications that will assist a permanent human presence in space.