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

Throughout this paper, we focus on the improvement of the direct feedbackalignment (DFA) algorithm and extend the usage of the DFA to convolutional andrecurrent neural networks (CNNs and RNNs). Even though the DFA algorithm isbiologically plausible and has a potential of high-speed training, it has notbeen considered as the substitute for back-propagation (BP) due to the lowaccuracy in the CNN and RNN training. In this work, we propose a new DFAalgorithm for BP-level accurate CNN and RNN training. Firstly, we divide thenetwork into several modules and apply the DFA algorithm within the module.Second, the DFA with the sparse backward weight is applied. It comes with aform of dilated convolution in the CNN case, and in a form of sparse matrixmultiplication in the RNN case. Additionally, the error propagation method ofCNN becomes simpler through the group convolution. Finally, hybrid DFAincreases the accuracy of the CNN and RNN training to the BP-level while takingadvantage of the parallelism and hardware efficiency of the DFA algorithm.

Extension of Direct Feedback Alignment to Convolutional and Recurrent Neural Network for Bio-plausible Deep Learning