Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Dynamic cluster formation using level set methods.

Andy M Yip1, Chris Ding, Tony F Chan

  • 1Department of Mathematics, National University of Singapore, 2, Science Drive 2, Singapore 117543, Singapore. matymha@nus.edu.sg

IEEE Transactions on Pattern Analysis and Machine Intelligence
|May 27, 2006
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Weighted Sparse Partial Least Squares With Joint Sample and Feature Selection for Integrating Multi-Omics Data.

IEEE transactions on computational biology and bioinformatics·2025
Same author

Contrastive independent subspace analysis network for multi-view spatial information extraction.

Neural networks : the official journal of the International Neural Network Society·2025
Same author

Tensorized Bipartite Graph Learning for Multi-View Clustering.

IEEE transactions on pattern analysis and machine intelligence·2022
Same author

Non-Greedy L21-Norm Maximization for Principal Component Analysis.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2021
Same author

Labeled-Robust Regression: Simultaneous Data Recovery and Classification.

IEEE transactions on cybernetics·2020
Same author

Revisiting L<sub>2,1</sub>-Norm Robustness With Vector Outlier Regularization.

IEEE transactions on neural networks and learning systems·2020
Same journal

HardFlow: Hard-Constrained Sampling for Flow-Matching Models Via Trajectory Optimization.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Industrial Brain: Self-Evolving Neuro-Symbolic Autonomy with Causal Resilience for Cyber-Physical Systems.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Adaptive Hardness-Driven Dictionary Distillation for Incomplete Streaming View Clustering.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Mixture of Global and Local Experts with Diffusion Transformer for Controllable Face Generation.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Task-KV: Task-aware KV Cache Optimization via Semantic Differentiation of Attention Heads.

IEEE transactions on pattern analysis and machine intelligence·2026
Same journal

Achieving Text-based Person Retrieval with Any Granularity.

IEEE transactions on pattern analysis and machine intelligence·2026
See all related articles

This study introduces cluster intensity functions (CIFs) for robust density-based clustering, especially when clusters overlap. CIFs effectively identify cluster centers and boundaries, outperforming traditional methods like DBSCAN.

Area of Science:

  • Data Mining
  • Machine Learning
  • Computational Statistics

Background:

  • Density-based clustering methods excel at identifying arbitrary cluster shapes without pre-specifying the number of clusters.
  • Overlapping clusters pose challenges for traditional methods, blurring cluster centers and boundaries.
  • Kernel density estimation can lead to oscillatory or oversmoothed density functions, complicating analysis.

Purpose of the Study:

  • To introduce a novel approach for robust density-based clustering, particularly for overlapping clusters.
  • To develop a method that clearly delineates cluster centers, boundaries, and data point membership even in dense regions.
  • To enhance the reliability of clustering algorithms by addressing limitations of kernel density estimation.

Main Methods:

Related Experiment Videos

  • Introduction of the Cluster Intensity Function (CIF) to characterize clusters.
  • Application of Level Set Methods and related techniques to process density distributions.
  • Clustering via bump hunting and valley seeking on the derived CIFs.

    • Main Results:

    • CIFs effectively reveal cluster centers and boundaries even when clusters are in close proximity.
    • The proposed CIF-based bump hunting and valley seeking demonstrate greater robustness than density-based methods.
    • Comparisons show advantages over existing methods like DBSCAN and traditional valley seeking.

    Conclusions:

    • Cluster Intensity Functions offer a more reliable method for density-based clustering, especially in challenging overlapping scenarios.
    • Level Set Methods effectively resolve issues associated with kernel density estimation in clustering.
    • The developed approach provides improved accuracy and interpretability for cluster analysis.