eCite Digital Repository

Similarity weighted ensembles for relocating models of rare events


D'Este, CE and Rahman, A, Similarity weighted ensembles for relocating models of rare events, MCS 2013: Multiple Classifier Systems, 15-17 May 2013, Nanjing, China, pp. 25-36. ISBN 978-3-642-38066-2 (2013) [Refereed Conference Paper]

Not available

Copyright Statement

Copyright 2013 Springer-Verlag Berlin Heidelberg

Official URL:

DOI: doi:10.1007/978-3-642-38067-9_3


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. A novel method is presented for combining classifiers trained on regions with known sensor data and predicting rare events in new regions, specifically the closure of shellfish farms. The proposed similarity weighted ensemble method demonstrates an average 10 fold improvement in accuracy over One Class classification and 3 fold improvement over rules hand-crafted by an expert.

Item Details

Item Type:Refereed Conference Paper
Keywords:learned models, rare events, sensor data, prediction, shellfish farming
Research Division:Information and Computing Sciences
Research Group:Artificial intelligence
Research Field:Knowledge representation and reasoning
Objective Division:Animal Production and Animal Primary Products
Objective Group:Fisheries - aquaculture
Objective Field:Aquaculture molluscs (excl. oysters)
UTAS Author:D'Este, CE (Dr Claire D'Este)
ID Code:116724
Year Published:2013
Deposited By:Information and Communication Technology
Deposited On:2017-05-17
Last Modified:2017-06-21

Repository Staff Only: item control page