Beyond 3-D: USU Data Scientist Introduces Novel AI Tool to Interpret Complex Biological Data
Beyond 3-D: USU Data Scientist Introduces Novel AI Tool to Interpret Complex Biological Data
Publish Date: 2026-06-30 17:03:00
Source Domain: www.usu.edu
- The article discusses the challenges in visualizing complex, high-dimensional data, such as biological processes, and how human-engineered scopes cannot capture these sub-visible structures.
- Kevin Moon, director of USU’s Data Science and Artificial Intelligence Center, introduces RF-PHATE, a supervised data visualization method that allows for the exploration of relevant data relationships in multi-dimensional datasets.
- RF-PHATE outperforms traditional unsupervised methods by better preserving the structure and interrelations of data groups.
- The team demonstrated RF-PHATE’s effectiveness by identifying a previously suspected sub-type in multiple sclerosis data, which can guide more tailored treatment decisions.
- RF-PHATE is not limited to biological data; it has wide applicability across various fields and can be used to develop more interpretable AI models and analyze AI systems themselves.
- The research, supported by the National Institutes of Health and the IVADO Visiting Scholar Program, represents an interdisciplinary effort involving multiple international institutions, showcasing collaborative efforts to boost scientific research through AI and machine learning.
- The paper emphasizes the importance of interdisciplinary collaborations to effectively analyze scientific data and foster discovery, encouraging students to explore this research area.
