Learning Deep Belief Networks from Non-Stationary Streams

File Description SizeFormat 
calandra_icann2012.pdfAccepted version1.33 MBAdobe PDFDownload
Title: Learning Deep Belief Networks from Non-Stationary Streams
Author(s): Calandra, R
Raiko, T
Deisenroth, MP
Pouzols, FM
Item Type: Conference Paper
Abstract: Deep learning has proven to be beneficial for complex tasks such as classifying images. However, this approach has been mostly applied to static datasets. The analysis of non-stationary (e.g., concept drift) streams of data involves specific issues connected with the temporal and changing nature of the data. In this paper, we propose a proof-of-concept method, called Adaptive Deep Belief Networks, of how deep learning can be generalized to learn online from changing streams of data. We do so by exploiting the generative properties of the model to incrementally re-train the Deep Belief Network whenever new data are collected. This approach eliminates the need to store past observations and, therefore, requires only constant memory consumption. Hence, our approach can be valuable for life-long learning from non-stationary data streams. © 2012 Springer-Verlag.
Publication Date: 25-Oct-2012
URI: http://hdl.handle.net/10044/1/12208
DOI: http://dx.doi.org/10.1007/978-3-642-33266-1_47
ISBN: 978-3-642-33265-4
ISSN: 0302-9743
Publisher: Springer Berlin Heidelberg
Start Page: 379
End Page: 386
Journal / Book Title: Artificial Neural Networks and Machine Learning – ICANN 2012
Volume: 7553
Copyright Statement: © 2012 Springer-Verlag Berlin Heidelberg. The original publication is available at www.springerlink.com
Conference Name: ICANN 2012
Start Date: 2012-09-11
Finish Date: 2012-09-14
Conference Place: Lausanne, Switzerland
Appears in Collections:Computing

Items in Spiral are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commons