Missing values: how many can they be to preserve classification reliability?
Martti JuholaJorma LaurikkalaPublished in: Artif. Intell. Rev. (2013)
Keyphrases
- missing values
- cost sensitive learning
- missing data
- data imputation
- multiple imputation
- feature values
- incomplete data
- missing data imputation
- incomplete data sets
- multivariate temporal data
- pattern recognition
- unseen test data
- support vector machine
- classification algorithm
- classification accuracy
- missing attribute values
- decision trees
- feature extraction
- high dimensional data
- machine learning
- data stream classification
- statistical databases
- feature vectors
- imputation methods
- unseen data
- probabilistic model
- cost sensitive
- data sets
- principal component analysis
- neural network