In Bilbao, the International Summer School on Deep Learning took place between 17 and 21 July 2017. The 1-week event gathered more than 1300 PhD students, researchers and AI practitioners with an interested in the theoretical and practical advances of deep learning at large. This international event featured many big names in the ever-more-popular deep learning world, but games had an emphatic presence!
IDG member Georgios N. Yannakakis gave three lectures on deep learning in games with a interesting twist: not for playing them. The lectures instead focused on the conjunction of deep learning and preference learning for modeling cognitive, affective and behavioral patterns of players as well as the use of deep learning and divergent search for the generation of game content in an autonomous way or with the assistance of a game designer. The various affect modeling and content generation projects he presented were based on some of the top-tier work of IDG's PhD and MSc students.