-Titile: Real-time Visual Analytics for User-Driven Machine Learning
Jaegul Choo (https://sites.google.com/site/jaegulchoo/ ) is an assistant professor in the Dept. of Computer Science and Engineering at Korea University. He has been a research scientist at Georgia Tech from 2011 to 2015, where we also received M.S in 2009 and Ph.D in 2013. His research focuses on visual analytics for high-dimensional data, which leverages both data mining and interactive visualization. He has been publishing in premier venues in both fields such as KDD, WWW, WSDM, ICDM, ICWSM, SDM, TVCG, CG&A, and VAST. He earned the Outstanding Research Scientist Award at Georgia Tech in 2015 and the Best Poster Award at IEEE VAST (as part of IEEE VIS) in 2014, and he was nominated as one of the four finalists in IEEE Visualization Pioneers Group dissertation award in 2013
Visual analytics, a recently emerging area, allows humans to explore big data via interactive visualization. One of the biggest challenges in visual analytics is how to utilize and improve computational methods, which reveals meaningful characteristics of large-scale data, in visual analytic environments so that they can properly serve end-users’ tasks and goals. In this talk, I will present mathematical and algorithmic approaches as well as visual analytics systems that address these issues. In detail, the talk will cover some of my research about interactive yet efficient techniques for topic modeling, 2D embedding, and deep learning, along with the visual analytics systems based on them.