IE Graduate Seminar: "Extreme Learning Machines for Visualization"

Thursday, February 2 at 3:30 pm to 4:20 pm
Seamans Center, 4030 SC
103 South Capitol Street, Iowa City, Iowa

"Extreme Learning Machines for Visualization," presented by Andrey Gritsenko, PhD Candidate, Industrial Engineering, University of Iowa Abstract: Visualization of data is an important, but not simple, task in the field of data analysis. Visualization of data in high-dimensional space can be described as a task of finding (or building) a new, low-dimensional (1D, 2D, or 3D), projection space, which allows to best represent the data. Also this task becomes even more complicated when we talk about Big Data problems. This talk will cover basics of the Extreme Learning Machines theory—a method that allows training of artificial neural networks in linear time. We will also discuss the implementation of this method for data visualization, applying the idea of reverse projection—ELMVIS. Further, the application of ELMVIS for classification and regression tasks will be presented, followed by the discussion on the influence of target information on visualization results. Individuals with disabilities are encouraged to attend all University of Iowa sponsored events. If you have a disability that requires an accommodation in order to participate in this program, please call the department in advance, at 319-335-5939.

Contact Info: MIE Office, mie-engineering@uiowa.edu, 319-335-5939