Neurocomputing Special Issue on Subspace Learning

Neurocomputing Special Issue on Subspace Learning

Subspace learning, a typical neurocomputing technique, sheds light on
various tasks from computer vision to data mining. A subspace learning
algorithm projects the original high dimensional feature space to a low
dimensional subspace, wherein specific statistical properties can be well
preserved. For example, Fisher’s linear discriminant analysis preserves the
classification structure in the low dimensional projected subspace by
assuming samples are drawn from homoscedastic unimodal Gaussians.

The past years witnessed very significant contributions of subspace learning
for various applications, e.g., face identification and authentication,
human gait analysis, video tracking, document classification, scene
understanding, object categorization, and multimedia information retrieval.
Subspace learning has been furthered from relevant research fields, e.g.,
nonparametric Bayesian analysis, manifold learning, spectral analysis,
kernel machine, tensor machine, and incremental learning. As a consequence,
subspace learning is a never‐ending resilience field and attracts growing
efforts from different fields.

Elsevier Neurocomputing hunts for original research results for a Special
Issue on Subspace Learning. The goals of this special issue are twofold: 1)
developing subspace learning algorithms to target specific applications and
2) defining applications, which can be cleared up by subspace learning
algorithms.

Manuscripts are solicited to address a wide range of topics in subspace
learning, but not limit to the following:
- Extensions of Fisher’s linear discriminant analysis and other traditional
subspace learning algorithms, e.g., principal component analysis.
- Spectral analysis
- Probabilistic graphical models
- Manifold learning
- Kernel machines and tensor machines
- Supervised, semi‐supervised and unsupervised subspace analysis
- Sparse analysis
- Subspace learning for biometrics, e.g., face.
- Subspace learning for multimedia information retrieval, e.g., relevance
feedback and active sample selection.
- Subspace learning for video surveillance, e.g., video tracking, motion
analysis and background modeling.

Manuscripts (6‐15 pages in the Neurocomputing publishing format) should be
submitted via the Electronic Editorial System, Elsevier:
http://ees.elsevier.com/neucom/

Important: when submitting your manuscript, at the step of “Selecting an
Article Type is Required for Submission”, please indicate: “Special Issue:
Subspace Learning”

Important Dates
Manuscript submission: 10 May 2009
Preliminary results: 10 July 2009
Revised version: 10 August 2009
Notification: 10 October 2009
Final manuscripts due: 10 November 2009
Anticipated publication: Spring 2010

Guest editors:
Xuelong Li
Birkbeck College, University of London
xuelong_li@ieee.org

Dacheng Tao
Nanyang Technological University
dacheng.tao@gmail.com

Received on Tuesday, 21 April 2009 06:16:59 UTC