(comming soon) Across-media patterns of media use and evolution.
Visualization as a spatial coordination game. Practical application of game-theory in an important problem, fast and low-stress embeddings.
User and Machine draw and edit sketches together. Learn modifications online and graphically (for websearches, drawings, diagrammatic command diagrams, etc.)
Models of how communities obtain information (and learn) cooperatively, and how they can improve. Causality-based (towards full-time emplyoment) activities recommendation system.
Game in the image grid, players correspond to different shape classes. Fast and accurate results. Super-parallel (CUDA) implementation.
Machine Learning problems as dynamical systems over graphs, attractors as players' solutions (MIT Thesis).
Graphical creatures (i.e., of masses and springs) learn to follow and control a ball together. Learning studied with different 'graphs' (body configurations) and algorithms.
I'm a University of Cambridge PhD and former MIT Medialab student and the founder of Contra-Ataque!, a Brazillian Non-profit Organization. I've worked with some pretty impressive people both there and the University of Tokyo.