NIPS
2004
13 years 6 months ago
2004
We describe a way of using multiple different types of similarity relationship to learn a low-dimensional embedding of a dataset. Our method chooses different, possibly overlappin...