Expressive power of recurrent neural networks

Valentin Khrulkov, Alexander Novikov, Ivan Oseledets

    Research output: Contribution to conferencePaperpeer-review

    25 Citations (Scopus)

    Abstract

    Deep neural networks are surprisingly efficient at solving practical tasks, but the theory behind this phenomenon is only starting to catch up with the practice. Numerous works show that depth is the key to this efficiency. A certain class of deep convolutional networks – namely those that correspond to the Hierarchical Tucker (HT) tensor decomposition – has been proven to have exponentially higher expressive power than shallow networks. I.e. a shallow network of exponential width is required to realize the same score function as computed by the deep architecture. In this paper, we prove the expressive power theorem (an exponential lower bound on the width of the equivalent shallow network) for a class of recurrent neural networks – ones that correspond to the Tensor Train (TT) decomposition. This means that even processing an image patch by patch with an RNN can be exponentially more efficient than a (shallow) convolutional network with one hidden layer. Using theoretical results on the relation between the tensor decompositions we compare expressive powers of the HT- and TT-Networks. We also implement the recurrent TT-Networks and provide numerical evidence of their expressivity.

    Original languageEnglish
    Publication statusPublished - 2018
    Event6th International Conference on Learning Representations, ICLR 2018 - Vancouver, Canada
    Duration: 30 Apr 20183 May 2018

    Conference

    Conference6th International Conference on Learning Representations, ICLR 2018
    Country/TerritoryCanada
    CityVancouver
    Period30/04/183/05/18

    Fingerprint

    Dive into the research topics of 'Expressive power of recurrent neural networks'. Together they form a unique fingerprint.

    Cite this