 |
 |
Science at the Edge November 3, 2017 1400 Biomedical Physical Sciences 11:30 am John P. Ralston Department of Physics and Astronomy University of Kansas "Finding Emergent Regularities in Multi-Dimensional Data: If Any!" Abstract: 21st century sciences faces the problem of too much (!) high-dimensional data to interpret or use effectively. It is a great interdisciplinary "crisis" of the current era. The problem is mathematically deep. It is also an unrecognized outcome of the reductionist approach to science exemplified by 19th century physics, which is obsolete. Seeking "emergent" subsystems is a new idea to discover patterns of orderly behavior that vast complexity often makes on its own. By abandoning reductionism, the critical concept of 'what do you mean by a probability' is freed up to allow progress. We will describe the outcomes, with real-world examples, where anyone can usually find a bit of gold in almost any data. Lerena R. Heintzelman Department of Physics & Astronomy Michigan State University 567 Wilson Rd. Room 3261 East Lansing, MI 48824 517-884-5513