Corpus ID: 60441374

PAC-Bayes Analysis of Sentence Representation

  title={PAC-Bayes Analysis of Sentence Representation},
  author={Kento Nozawa and I. Sato},
  • Kento Nozawa, I. Sato
  • Published 2019
  • Computer Science, Mathematics
  • ArXiv
  • Learning sentence vectors from an unlabeled corpus has attracted attention because such vectors can represent sentences in a lower dimensional and continuous space. Simple heuristics using pre-trained word vectors are widely applied to machine learning tasks. However, they are not well understood from a theoretical perspective. We analyze learning sentence vectors from a transfer learning perspective by using a PAC-Bayes bound that enables us to understand existing heuristics. We show that… CONTINUE READING


    Supervised Learning of Universal Sentence Representations from Natural Language Inference Data
    • 1,046
    • PDF
    Efficient Estimation of Word Representations in Vector Space
    • 15,724
    • Highly Influential
    • PDF
    Enriching Word Vectors with Subword Information
    • 3,972
    • PDF
    Advances in Pre-Training Distributed Word Representations
    • 484
    • PDF
    Learning Word Vectors for Sentiment Analysis
    • 2,028
    • Highly Influential
    • PDF
    Natural Language Processing (Almost) from Scratch
    • 5,637
    • PDF
    Distributed Representations of Words and Phrases and their Compositionality
    • 19,642
    • Highly Influential
    • PDF
    Glove: Global Vectors for Word Representation
    • 15,494
    • PDF
    Learning Word Vectors for 157 Languages
    • 482
    • PDF