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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Gated Recurrent Unit (GRU) is a recently-developed variation of the longshort-term memory (LSTM) unit, both of which are types of recurrent neuralnetwork (RNN). Through empirical evidence, both models have been proven to beeffective in a wide variety of machine learning tasks such as natural languageprocessing (Wen et al., 2015), speech recognition (Chorowski et al., 2015), andtext classification (Yang et al., 2016). Conventionally, like most neuralnetworks, both of the aforementioned RNN variants employ the Softmax functionas its final output layer for its prediction, and the cross-entropy functionfor computing its loss. In this paper, we present an amendment to this norm byintroducing linear support vector machine (SVM) as the replacement for Softmaxin the final output layer of a GRU model. Furthermore, the cross-entropyfunction shall be replaced with a margin-based function. While there have beensimilar studies (Alalshekmubarak & Smith, 2013; Tang, 2013), this proposal isprimarily intended for binary classification on intrusion detection using the2013 network traffic data from the honeypot systems of Kyoto University.Results show that the GRU-SVM model performs relatively higher than theconventional GRU-Softmax model. The proposed model reached a training accuracyof ~81.54% and a testing accuracy of ~84.15%, while the latter was able toreach a training accuracy of ~63.07% and a testing accuracy of ~70.75%. Inaddition, the juxtaposition of these two final output layers indicate that theSVM would outperform Softmax in prediction time - a theoretical implicationwhich was supported by the actual training and testing time in the study.

A Neural Network Architecture Combining Gated Recurrent Unit (GRU) and Support Vector Machine (SVM) for Intrusion Detection in Network Traffic Data