Intrinsic dimension of data representations in deep neural networks.
Alessio AnsuiniAlessandro LaioJakob H. MackeDavide ZoccolanPublished in: NeurIPS (2019)
Keyphrases
- data representations
- intrinsic dimension
- neural network
- data representation
- similarity search
- intrinsic dimensionality
- pattern recognition
- high dimensions
- dimensionality reduction
- distance function
- high dimensional data
- nearest neighbor
- dimension reduction
- diffusion maps
- query processing
- high dimensional
- learning algorithm
- xml documents
- low dimensional
- multi dimensional
- data distribution
- databases
- binary codes
- nearest neighbor search
- noisy data
- sparse representation
- similarity measure
- probabilistic model