Multidimensional Projection at Scale on the Web: Tips and Tricks

April 11, 2019

Time: April 11, 2019
Download as iCal:

In the recent years, multidimensional projection has become more popular thanks to algorithms such as t-SNE and UMAP. New articles have been published to explain tasks and issues related to the multiple algorithms at hand, and scalability has improved dramatically.

In this talk, I will present some of our recent work to improve the scalability of multidimensional projections, in particular stressing all the details to address to reach the visualization of several millions of projected points online. The talk will be a conversation about the methods we use and potential improvements given novel algorithms and data structures that become usable, such as compressed bitmaps for dynamic queries, fast approximate nearest-neighbors, and novel algorithms such as UMAP that offer interesting opportunities for scalability. I will also raise some challenges that remain to be solved to make multidimensional projections usable for a larger audience.

 

Biography: 

Jean-Daniel Fekete is Senior Research Scientist (DR1) at INRIA, Scientific Leader of the INRIA Project Team AVIZ that he founded in 2007.

He received his PhD in Computer Science in 1996 from Université Paris-Sud. From 1997 to 2001, he joined the Graphic Design group at the Ecole des Mines de Nantes that he led from 2000 to 2001. He was then invited to join the Human-Computer Interaction Laboratory at the University of Maryland in the USA in 2001-2002. He was recruited by INRIA in 2002 as a confirmed researcher and became Senior Research Scientist in 2006. In 2015, he was on Sabbatical at the Visualization and Computer Graphics group at NYU-Poly, and at the Visual Computing Group at Harvard.

His main Research areas are Visual Analytics, Information Visualization and Human Computer Interaction. He published more than 150 articles in multiple conferences and journals.

He is member of the Eurographics publication board, and Associate Editor-in-Chief of IEEE Transactions on Visualization and Computer Graphics.

URL:  http://www.aviz.fr/~fekete

To the top of the page