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

Unsupervised structure learning in high-dimensional time series data hasattracted a lot of research interests. For example, segmenting and labellinghigh dimensional time series can be helpful in behavior understanding andmedical diagnosis. Recent advances in generative sequential modeling havesuggested to combine recurrent neural networks with state space models (e.g.,Hidden Markov Models). This combination can model not only the long termdependency in sequential data, but also the uncertainty included in the hiddenstates. Inheriting these advantages of stochastic neural sequential models, wepropose a structured and stochastic sequential neural network, which modelsboth the long-term dependencies via recurrent neural networks and theuncertainty in the segmentation and labels via discrete random variables. Foraccurate and efficient inference, we present a bi-directional inference networkby reparamterizing the categorical segmentation and labels with the recentproposed Gumbel-Softmax approximation and resort to the Stochastic GradientVariational Bayes. We evaluate the proposed model in a number of tasks,including speech modeling, automatic segmentation and labeling in behaviorunderstanding, and sequential multi-objects recognition. Experimental resultshave demonstrated that our proposed model can achieve significant improvementover the state-of-the-art methods.

Stochastic Sequential Neural Networks with Structured Inference