Dimensionality Selection of Hyperbolic Graph Embeddings using Decomposed Normalized Maximum Likelihood Code-Length.
Ryo YukiYuichi IkeKenji YamanishiPublished in: ICDM (2022)
Keyphrases
- normalized maximum likelihood
- code length
- change point detection
- gray code
- dimensionality reduction
- spatially adaptive
- minimum description length
- feature space
- high dimensional
- low dimensional
- binary codes
- channel capacity
- channel coding
- high dimensional data
- mdl principle
- feature selection
- locally adaptive
- sequential data
- multiresolution
- computational complexity