Establishing strong imputation performance of a denoising autoencoder in a wide range of missing data problems.
Najmeh AbiriBjörn LinsePatrik EdénMattias OhlssonPublished in: Neurocomputing (2019)
Keyphrases
- missing data
- missing values
- wide range
- denoising
- low rank
- incomplete data
- structure from motion
- data imputation
- multiple imputation
- motion segmentation
- perfect phylogeny
- image denoising
- matrix factorization
- neural network
- matrix completion
- dirichlet process mixture models
- missing information
- markov random field
- probabilistic model
- feature extraction