PLOTCON 2017: Aneesh Karve, Visualization without Guesswork смотреть онлайн

Until recently, guidelines for visualization design have been highly subjective, focusing on maxims, best practices, examples, and expert opinions. Although useful in practice, subjective design guidelines often lead to distortions that misrepresent the underlying data. Recent developments in mathematical models of visualization are now making it possible to design clear, accurate visualizations that communicate data at high fidelity. As a result, visualization design becomes more objective and visualizations can be automatically checked by computers. In this talk we’ll demonstrate how anyone can apply the principles of algebraic visualization to design more effective visualizations. We’ll further use Jupyter to check visualizations for semantic “bugs” like confusion, hallucination, jumblers, and misleaders.

Biography
Aneesh Karve is co-founder and Chief Technology Officer at Quilt Data. He has worked as a product manager, lead designer, and software engineer at companies like Microsoft, NVIDIA, and Matterport. Aneesh was the general manager and founding member of AdJitsu, the first realtime 3D advertising platform for iOS, acquired by Amobee in 2012.
Aneesh is currently advancing an open source data compiler and package manager that brings the power of source code management to big data. Recently, Aneesh has been applying mathematics to detect and neutralize bias in artificial intelligence and visualization (Strata 2016).
Aneesh holds degrees in chemistry, mathematics, and computer science. His research background spans machine learning, abstract algebra, and information visualization.