English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The interpretability of deep learning models has raised extended attentionthese years. It will be beneficial if we can learn an interpretable structurefrom deep learning models. In this paper, we focus on Recurrent NeuralNetworks~(RNNs) especially gated RNNs whose inner mechanism is still notclearly understood. We find that Finite State Automaton~(FSA) that processessequential data has more interpretable inner mechanism according to thedefinition of interpretability and can be learned from RNNs as theinterpretable structure. We propose two methods to learn FSA from RNN based ontwo different clustering methods. With the learned FSA and via experiments onartificial and real datasets, we find that FSA is more trustable than the RNNfrom which it learned, which gives FSA a chance to substitute RNNs inapplications involving humans' lives or dangerous facilities. Besides, weanalyze how the number of gates affects the performance of RNN. Our resultsuggests that gate in RNN is important but the less the better, which could bea guidance to design other RNNs. Finally, we observe that the FSA learned fromRNN gives semantic aggregated states and its transition graph shows us a veryinteresting vision of how RNNs intrinsically handle text classification tasks.

Learning with Interpretable Structure from Gated RNN