Using Feature Grouping as a Stochastic Regularizer for High-Dimensional Noisy Data.
Sergül AydöreBertrand ThirionOlivier GriselGaël VaroquauxPublished in: CoRR (2018)
Keyphrases
- noisy data
- high dimensional
- intrinsic dimensionality
- learning from noisy data
- noise tolerant
- total variation
- high dimensionality
- low dimensional
- similarity search
- image features
- noise free
- feature vectors
- nearest neighbor
- high dimensional data
- semi supervised
- missing data
- data points
- dimensionality reduction
- multi dimensional
- input data
- training data
- learning algorithm
- image data
- feature ranking
- feature space