An adaptive SIR method for block-wise evolving data streams

Chavent, Marie, Girard, Stéphane, Kuentz, Vanessa, Liquet, Benoît, Nguyen, Thi Mong Ngoc and Saracco, Jérôme (2011). An adaptive SIR method for block-wise evolving data streams. In: Proceedings of The 14th Conference of the ASMDA (Applied Stochastic Models and Data Analysis) International Society (ASMDA 2011). ASMDA 2011: XIVth International Symposium of Applied Stochastic Models and Data Analysis, Rome, Italy, (). 7-10 June, 2011.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
Author Chavent, Marie
Girard, Stéphane
Kuentz, Vanessa
Liquet, Benoît
Nguyen, Thi Mong Ngoc
Saracco, Jérôme
Title of paper An adaptive SIR method for block-wise evolving data streams
Conference name ASMDA 2011: XIVth International Symposium of Applied Stochastic Models and Data Analysis
Conference location Rome, Italy
Conference dates 7-10 June, 2011
Proceedings title Proceedings of The 14th Conference of the ASMDA (Applied Stochastic Models and Data Analysis) International Society (ASMDA 2011)
Place of Publication Chania, Crete, Greece
Publisher ASMDA International Society
Publication Year 2011
Sub-type Fully published paper
Open Access Status
Total pages 8
Language eng
Formatted Abstract/Summary
In this communication, we consider block-wise evolving data streams. When a semiparametric regression model involving a common dimension reduction direction β is assumed for each block, we propose an adaptive SIR (for sliced inverse regression) estimator of β. This estimator is faster than usual SIR applied to the union of all the blocks, both from computational complexity and running time points of view. We show the consistency of our estimator at the root-𝑛 rate. In a simulation, we illustrate the good numerical behaviour of the estimator. We also provide a graphical tool in order to detect if there exists a drift of the dimension reduction direction or some aberrant blocks of data. We illustrate our approach with various scenarios. Finally, possible extensions of this method are given.
Keyword Dimension reduction
Sliced inverse regression (SIR)
Data stream
Q-Index Code EX
Q-Index Status Provisional Code
Institutional Status Non-UQ

Document type: Conference Paper
Collection: School of Mathematics and Physics
 
Versions
Version Filter Type
Citation counts: Google Scholar Search Google Scholar
Created: Tue, 23 Sep 2014, 12:43:30 EST by Jon Swabey on behalf of School of Mathematics & Physics