This infrastructure award is for the development of a visualization and machine learning laboratory. The laboratory consists of powerful graphics workstations, a parallel processor, a large file server, a fiber optics high speed data network, and support personnel to maintain the laboratory. The research supported by this equipment includes: statistical data analysis, pattern recognition, and machine learning; visualization of sample volume data; interactive steering of simulations; high speed switching;concurrent systems; and applications of electronic libraries. A major bottleneck to the utilization of scientific information is the massive amount of data now collected. Scientific visualization is the term that covers an emerging collection of new techniques for transforming multidimensional numeric data into understandable graphic images. These images can then be manipulated and explored by scientists and engineers who, sitting at their workstations, create models and draw inferences from them. Machine learning includes a wide variety of techniques from the fields of pattern recognition, signal processing, statistical decision and control theory, and artificial intelligence that allow the machine itself to participate in the process of creating models and drawing inferences from data. It is anticipated that a new synergy between machine learning and scientific visualization will arise from this award and that substantial collaborations with scientists outside of the computer science department will occur.
