Computational Statistics and Visualization
Departmental researchers in the fields of computational statistics and visualization develop new statistical algorithms and software and advance statistical visualization methods. They work with data sets that are getting larger and more heterogeneous from year to year. They also work with previously unknown types of data obtained from new medical, scientific, and technological sources. These researchers extensively collaborate (and look for new data) with other researchers from across the university, the United States, and worldwide in medicine, biology, genetics, natural resources, environmental sciences, kinesiology, education, etc.
Some of the techniques developed and advanced by researchers in the department include random forests, linked micromap plots (a series of small maps that enhance traditional statistical maps), and archetypal analysis (a method that describes each individual in a data set as a mixture of pure individual types). All kinds of data from the web (in html, XML, or JSON format), technology (such as actigraph and eye-tracking devices), and in a geographic context (such as ESRI shapefiles and MISR data from the Jet Propulsion Lab) have been used in research projects from these researchers and can also be found in several of their classes. In this area, novel computational and visualization tools are valued even if traditional statistical theory is lacking.