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

Dyadic prediction methods operate on pairs of objects (dyads), aiming toinfer labels for out-of-sample dyads. We consider the full and almost full coldstart problem in dyadic prediction, a setting that occurs when both objects inan out-of-sample dyad have not been observed during training, or if one of themhas been observed, but very few times. A popular approach for addressing thisproblem is to train a model that makes predictions based on a pairwise featurerepresentation of the dyads, or, in case of kernel methods, based on a tensorproduct pairwise kernel. As an alternative to such a kernel approach, weintroduce a novel two-step learning algorithm that borrows ideas from thefields of pairwise learning and spectral filtering. We show theoretically thatthe two-step method is very closely related to the tensor product kernelapproach, and experimentally that it yields a slightly better predictiveperformance. Moreover, unlike existing tensor product kernel methods, thetwo-step method allows closed-form solutions for training and parameterselection via cross-validation estimates both in the full and almost full coldstart settings, making the approach much more efficient and straightforward toimplement.

A two-step learning approach for solving full and almost full cold start problems in dyadic prediction