Skip to main content

Relating RNN layers with the spectral WFA ranks in sequence modelling

Liza, Farhana Ferdousi, Grzes, Marek (2019) Relating RNN layers with the spectral WFA ranks in sequence modelling. In: ACL workshop on Deep Learning and Formal Languages: Building Bridges, 2 August 2019, Florence, Italy. (In press)

PDF - Pre-print
Download (373kB) Preview
[img]
Preview

Abstract

We analyse Recurrent Neural Networks (RNNs) to understand the significance of multiple LSTM layers. We argue that the Weighted Finite-state Automata (WFA) trained using a spectral learning algorithm are helpful to analyse RNNs. Our results suggest that multiple LSTM layers in RNNs help learning distributed hidden states, but have a smaller impact on the ability to learn long-term dependencies. The analysis is based on the empirical results, however relevant theory (whenever possible) was discussed to justify and support our conclusions.

Item Type: Conference or workshop item (Speech)
Uncontrolled keywords: deep learning, NLP, LSTM, WFA, spectral learning
Subjects: Q Science > QA Mathematics (inc Computing science) > QA 76 Software, computer programming, > QA76.76 Computer software
Divisions: Faculties > Sciences > School of Computing > Computational Intelligence Group
Depositing User: Marek Grzes
Date Deposited: 04 Jun 2019 11:14 UTC
Last Modified: 08 Jul 2019 08:36 UTC
Resource URI: https://kar.kent.ac.uk/id/eprint/74240 (The current URI for this page, for reference purposes)
  • Depositors only (login required):

Downloads

Downloads per month over past year