Analyzing High-Dimensional Data by Subspace Validity

13 years 9 months ago

Download infovis.uni-konstanz.de

We are proposing a novel method that makes it possible to analyze high dimensional data with arbitrary shaped projected clusters and high noise levels. At the core of our method lies the idea of subspace validity. We map the data in a way that allows us to test the quality of subspaces using statistical tests. Experimental results, both on synthetic and real data sets, demonstrate the potential of our method.

Amihood Amir, Reuven Kashi, Nathan S. Netanyahu, D

Real-time Traffic

Data Mining | ICDM 2003 | Real Data Sets | Subspace Validity |

claim paper

» Nearest Neighbor Search on Vertically Partitioned HighDimensional Data

» Unsupervised Feature Selection in High Dimensional Spaces and Uncertainty

» Subspace Analysis Using Random Mixture Models

» Local Adaptive Subspace Regression

» A Subspace Kernel for Nonlinear Feature Extraction

» A Linear Subspace Learning Approach via Sparse Coding

» Failure detection and localization in component based systems by online tracking

» Optimization of SelfOrganizing Maps Ensemble in Prediction

Post Info
More Details (n/a)

Added	04 Jul 2010
Updated	04 Jul 2010
Type	Conference
Year	2003
Where	ICDM
Authors	Amihood Amir, Reuven Kashi, Nathan S. Netanyahu, Daniel A. Keim, Markus Wawryniuk

Comments (0)

Sciweavers

Analyzing High-Dimensional Data by Subspace Validity

Data Mining | ICDM 2003 | Real Data Sets | Subspace Validity |

Explore & Download

Productivity Tools

Document Tools

Image Tools

Sciweavers