Paper Search Console

Home Search Page About Contact

Journal Title

Title of Journal: Pattern Anal Applic

Search In Journal Title:

Abbravation: Pattern Analysis and Applications

Search In Journal Abbravation:

Publisher

Springer-Verlag

Search In Publisher:

DOI

10.1007/bf01414500

Search In DOI:

ISSN

1433-755X

Search In ISSN:
Search In Title Of Papers:

Unsupervised learning of arbitrarily shaped cluste

Authors: Hichem Frigui
Publish Date: 2005/08/09
Volume: 8, Issue: 1-2, Pages: 32-49
PDF Link

Abstract

We propose a new clustering algorithm called SyMP which is based on synchronization of pulsecoupled oscillators SyMP represents each data point by an IntegrateandFire oscillator and uses the relative similarity between the points to model the interaction between the oscillators SyMP is robust to noise and outliers determines the number of clusters in an unsupervised manner and identifies clusters of arbitrary shapes The robustness of SyMP is an intrinsic property of the synchronization mechanism To determine the optimum number of clusters SyMP uses a dynamic and cluster dependent resolution parameter To identify clusters of various shapes SyMP models each cluster by an ensemble of Gaussian components SyMP does not require the specification of the number of components for each cluster This number is automatically determined using a dynamic intracluster resolution parameter Clusters with simple shapes would be modeled by few components while clusters with more complex shapes would require a larger number of components The proposed clustering approach is empirically evaluated with several synthetic data sets and its performance is compared with GK and CURE To illustrate the performance of SyMP on real and highdimensional data sets we use it to categorize two image databasesThe author would like to thank Dr Han from the Dept of Computer Science Eng Univ of Minnesota for providing the code for the CURE algorithm and Dr Boujemaa from the IMEDIA research group at INRIA France for providing the image database This material is based upon work supported by the National Science Foundation under award No IIS0133415

Keywords:

References

.
Search In Abstract Of Papers:
Other Papers In This Journal:
  1. Texture code matrix-based multi-instance iris recognition
  2. Wavelet feature space in computer-aided electroretinogram evaluation
  3. Computer vision techniques for traffic flow computation
  4. Computer vision techniques for traffic flow computation
  5. A systematic evaluation of the scale invariance of texture recognition methods
  6. Supervised texture classification: color space or texture feature selection?
  7. Multi-algorithmic cancelable fingerprint template generation based on weighted sum rule and T -operators
  8. Classification of dissimilarity data with a new flexible Mahalanobis-like metric
  9. Topographic Bernoulli block mixture mapping for binary tables
  10. A modified kernel clustering method with multiple factors
  11. Clustering spatial data with a hybrid EM approach
  12. The impact of false alarm mitigation on surface landmine detection in MWIR imagery
  13. On size invariance texture image retrieval by fuzzy logic classifier and scattering statistical features
  14. Automatic logo transition detection in digital video contents
  15. Parametric and nonparametric tests for speckled imagery
Search Result: