Cuckoo feature hashing: Dynamic weight sharing for sparse analytics

Jinyang Gao, Beng Chin Ooi, Yanyan Shen, Wang Chien Lee

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Scopus citations

Abstract

Feature hashing is widely used to process large scale sparse features for learning of predictive models. Collisions inherently happen in the hashing process and hurt the model performance. In this paper, we develop a new feature hashing scheme called Cuckoo Feature Hashing (CCFH), which treats feature hashing as a problem of dynamic weight sharing during model training. By leveraging a set of indicators to dynamically decide the weight of each feature based on alternative hash locations, CCFH effectively prevents the collisions between important features to the model, i.e. predictive features, and thus avoid model performance degradation. Experimental results on prediction tasks with hundred-millions of features demonstrate that CCFH can achieve the same level of performance by using only 15%-25% parameters compared with conventional feature hashing.

Original languageEnglish (US)
Title of host publicationProceedings of the 27th International Joint Conference on Artificial Intelligence, IJCAI 2018
EditorsJerome Lang
PublisherInternational Joint Conferences on Artificial Intelligence
Pages2135-2141
Number of pages7
ISBN (Electronic)9780999241127
DOIs
StatePublished - 2018
Event27th International Joint Conference on Artificial Intelligence, IJCAI 2018 - Stockholm, Sweden
Duration: Jul 13 2018Jul 19 2018

Publication series

NameIJCAI International Joint Conference on Artificial Intelligence
Volume2018-July
ISSN (Print)1045-0823

Other

Other27th International Joint Conference on Artificial Intelligence, IJCAI 2018
CountrySweden
CityStockholm
Period7/13/187/19/18

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence

Fingerprint Dive into the research topics of 'Cuckoo feature hashing: Dynamic weight sharing for sparse analytics'. Together they form a unique fingerprint.

Cite this