Diffusion K-means clustering on manifolds: provable exact recovery via semidefinite relaxations

@article{Chen2019DiffusionKC,
  title={Diffusion K-means clustering on manifolds: provable exact recovery via semidefinite relaxations},
  author={Xiaohui Chen and Y. Yang},
  journal={ArXiv},
  year={2019},
  volume={abs/1903.04416}
}
We introduce the {\it diffusion $K$-means} clustering method on Riemannian submanifolds, which maximizes the within-cluster connectedness based on the diffusion distance. The diffusion $K$-means constructs a random walk on the similarity graph with vertices as data points randomly sampled on the manifolds and edges as similarities given by a kernel that captures the local geometry of manifolds. The diffusion $K$-means is a multi-scale clustering tool that is suitable for data with non-linear… Expand
Nonlinear dimensionality reduction for clustering
Near-Optimal Comparison Based Clustering
Limitations of Big Data Partitions Technology
Cutoff for Exact Recovery of Gaussian Mixture Models
Grassmannian diffusion maps based dimension reduction and classification for high-dimensional data

References

SHOWING 1-10 OF 49 REFERENCES
Approximating K-means-type Clustering via Semidefinite Programming
Certifying Global Optimality of Graph Cuts via Semidefinite Relaxation: A Performance Guarantee for Spectral Clustering
Geometric structure of graph Laplacian embeddings
Learning by Unsupervised Nonlinear Diffusion
Sparse Manifold Clustering and Embedding
Path-Based Spectral Clustering: Guarantees, Robustness to Outliers, and Fast Algorithms
Clustering Based on Pairwise Distances When the Data is of Mixed Dimensions
  • E. Arias-Castro
  • Computer Science, Mathematics
  • IEEE Transactions on Information Theory
  • 2011
Fundamental Limitations of Spectral Clustering
The geometry of kernelized spectral clustering
PECOK: a convex optimization approach to variable clustering
...
1
2
3
4
5
...