A representation-learning game for classes of prediction tasks

arXiv:2403.06971v1 Announce Type: new
Abstract: We propose a game-based formulation for learning dimensionality-reducing representations of feature vectors, when only a prior knowledge on future prediction tasks is available. In this game, the first player chooses a representation, and then the second player adversarially chooses a prediction task from a given class, representing the prior knowledge. The first player aims is to minimize, and the second player to maximize, the regret: The minimal prediction loss using the representation, compared to the …

abstract arxiv class cs.it cs.lg dimensionality feature future game knowledge math.it prediction prior representation tasks type vectors