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 learning techniques have been applied widely in industrialrecommendation systems. However, far less attention has been paid to theoverfitting problem of models in recommendation systems, which, on thecontrary, is recognized as a critical issue for deep neural networks. In thecontext of Click-Through Rate (CTR) prediction, we observe an interestingone-epoch overfitting problem: the model performance exhibits a dramaticdegradation at the beginning of the second epoch. Such a phenomenon has beenwitnessed widely in real-world applications of CTR models. Thereby, the bestperformance is usually achieved by training with only one epoch. To understandthe underlying factors behind the one-epoch phenomenon, we conduct extensiveexperiments on the production data set collected from the display advertisingsystem of Alibaba. The results show that the model structure, the optimizationalgorithm with a fast convergence rate, and the feature sparsity are closelyrelated to the one-epoch phenomenon. We also provide a likely hypothesis forexplaining such a phenomenon and conduct a set of proof-of-concept experiments.We hope this work can shed light on future research on training more epochs forbetter performance.

Towards Understanding the Overfitting Phenomenon of Deep Click-Through Rate Prediction Models