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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Deep networks are an integral part of the current machine learning paradigm.Their inherent ability to learn complex functional mappings between data andvarious target variables, while discovering hidden, task-driven features, makesthem a powerful technology in a wide variety of applications. Nonetheless, thesuccess of these networks typically relies on the availability of sufficienttraining data to optimize a large number of free parameters while avoidingoverfitting, especially for networks with large capacity. In scenarios withlimited training budgets, e.g., supervised tasks with limited labeled samples,several generic and/or task-specific regularization techniques, including dataaugmentation, have been applied to improve the generalization of deepnetworks.Typically such regularizations are introduced independently of thatdata or training scenario, and must therefore be tuned, tested, and modified tomeet the needs of a particular network. In this paper, we propose a novelregularization framework that is driven by the population-level statistics ofthe feature space to be learned. The regularization is in the form of a\textbf{cooperating subnetwork}, which is an auto-encoder architecture attachedto the feature space and trained in conjunction with the primary network. Weintroduce the architecture and training methodology and demonstrate theeffectiveness of the proposed cooperative network-based regularization in avariety of tasks and architectures from the literature. Our code is freelyavailable at \url{https://github.com/riddhishb/CoopSubNet

CoopSubNet: Cooperating Subnetwork for Data-Driven Regularization of Deep Networks under Limited Training Budgets