Using locally estimated geodesic distance to optimize neighborhood graph for isometric data embedding.
Guihua WenLijun JiangJun WenPublished in: Pattern Recognit. (2008)
Keyphrases
- geodesic distance
- manifold learning
- neighborhood graph
- euclidean space
- dimensionality reduction
- low dimensional
- high dimensional
- semi supervised
- feature extraction
- dimension reduction
- high dimensional data
- sparse representation
- euclidean distance
- data points
- least squares
- geometric structure
- multidimensional scaling
- maximum likelihood
- unsupervised learning
- shortest path
- similarity search
- vector space
- image analysis
- similarity measure
- clustering algorithm