eCite Digital Repository

Ensemble aggregation methods for relocating models of rare events


D'Este, C and Timms, G and Turnbull, A and Rahman, A, Ensemble aggregation methods for relocating models of rare events, Engineering Applications of Artificial Intelligence, 34 pp. 58-65. ISSN 0952-1976 (2014) [Refereed Article]

Copyright Statement

Crown Copyright 2014 Published by Elsevier Ltd.

DOI: doi:10.1016/j.engappai.2014.05.007


Spatially distributed regions may have different influences that affect the underlying physical processes and make it inappropriate to directly relocate learned models. We may also be aiming to detect rare events for which we have examples in some regions, but not others. Three novel voting methods are presented for combining classifiers trained on regions with available examples for predicting rare events in new regions; specifically the closure of shellfish farms. The ensemble methods introduced are consistently more accurate at predicting closures. Approximately 63% of locations were successfully learned with Class Balance aggregation compared with 37% for the Expert guidelines, and 0% for One Class Classification.

Item Details

Item Type:Refereed Article
Keywords:aquaculture, ensemble classifiers, rare event detection
Research Division:Agricultural, Veterinary and Food Sciences
Research Group:Fisheries sciences
Research Field:Aquaculture
Objective Division:Animal Production and Animal Primary Products
Objective Group:Fisheries - aquaculture
Objective Field:Aquaculture crustaceans (excl. rock lobster and prawns)
UTAS Author:D'Este, C (Dr Claire D'Este)
UTAS Author:Timms, G (Dr Gregory Timms)
ID Code:118070
Year Published:2014
Web of Science® Times Cited:7
Deposited By:Information and Communication Technology
Deposited On:2017-07-03
Last Modified:2017-10-16

Repository Staff Only: item control page