Disentangling Identifiable Features from Noisy Data with Structured Nonlinear ICA.
Hermanni HälväSylvain Le CorffLuc LehéricyJonathan SoYongjie ZhuElisabeth GassiatAapo HyvärinenPublished in: CoRR (2021)
Keyphrases
- noisy data
- independent component analysis
- high dimensionality
- noise tolerant
- noise free
- feature extraction
- missing values
- missing data
- computer vision
- linear filters
- feature set
- classification accuracy
- high dimensional
- image processing
- input data
- signal processing
- denoising
- decision trees
- intrinsic dimensionality
- data mining