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

Deep neural network (DNN) based approaches hold significant potential forreinforcement learning (RL) and have already shown remarkable gains overstate-of-art methods in a number of applications. The effectiveness of DNNmethods can be attributed to leveraging the abundance of supervised data tolearn value functions, Q-functions, and policy function approximations withoutthe need for feature engineering. Nevertheless, the deployment of DNN-basedpredictors with very deep architectures can pose an issue due to computationaland other resource constraints at test-time in a number of applications. Wepropose a novel approach for reducing the average latency by learning acomputationally efficient gating function that is capable of recognizing statesin a sequential decision process for which policy prescriptions of a shallownetwork suffices and deeper layers of the DNN have little marginal utility. Theoverall system is adaptive in that it dynamically switches control actionsbased on state-estimates in order to reduce average latency without sacrificingterminal performance. We experiment with a number of alternative loss-functionsto train gating functions and shallow policies and show that in a number ofapplications a speed-up of up to almost 5X can be obtained with little loss inperformance.

Sequential Dynamic Decision Making with Deep Neural Nets on a Test-Time Budget