A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms

    Tristan Deleu
    Tristan Deleu
    Nasim Rahaman
    Nasim Rahaman
    Nan Rosemary Ke
    Nan Rosemary Ke
    Sebastien Lachapelle
    Sebastien Lachapelle
    Olexa Bilaniuk
    Olexa Bilaniuk
    被引用次数29|引用|412|

    international conference on learning representations, 2020.

    摘要:

    We propose to use a meta-learning objective that maximizes the speed of transfer on a modified distribution to learn how to modularize acquired knowledge. In particular, we focus on how to factor a joint distribution into appropriate conditionals, consistent with the causal directions. We explain when this can work, using the assumption t...更多
    您的评分 :
    0

     

    标签
    评论