Clustering for directed graphs using parametrized random walk diffusion kernels.
Harry SeviMatthieu JonckheereArgyris KalogeratosPublished in: CoRR (2022)
Keyphrases
- random walk
- directed graph
- proximity measures
- spectral methods
- markov chain
- markov random walk
- heat kernel
- link prediction
- clustering method
- graph laplacian
- clustering algorithm
- k means
- transition probabilities
- strongly connected
- graph structure
- directed acyclic graph
- undirected graph
- kernel function
- search algorithm
- transition probability matrix
- data clustering
- flow graph
- graph structures
- spectral clustering
- machine learning
- kernel methods
- distance metric
- binary images
- feature space
- optimal solution
- learning algorithm