A Preconditioned Richardson Regularization for the Data Completion Problem and the Kozlov-Maz'ya-Fomin Method.
Duc Thang DuFaten JelassiPublished in: ARIMA J. (2010)
Keyphrases
- synthetic data
- test data
- prior information
- high accuracy
- missing values
- data analysis
- data collection
- prior knowledge
- input data
- similarity measure
- noisy data
- missing data
- training samples
- data sets
- feature set
- correlation analysis
- information loss
- high quality
- objective function
- preprocessing
- detection method
- clustering method
- training set
- pairwise
- significant improvement
- data streams
- cross validation
- em algorithm
- statistical methods
- training data
- k means