An exploratory analysis of data noisy scenarios in a Pareto-front based dynamic feature selection method.
Jhoseph JesusAnne M. P. CanutoDaniel AraújoPublished in: Appl. Soft Comput. (2021)
Keyphrases
- synthetic data
- noisy data
- feature selection
- input data
- mutual information
- statistical methods
- data sets
- database
- test data
- data structure
- informative features
- data quality
- prior information
- preprocessing
- data analysis
- missing data
- missing values
- clustering method
- irrelevant features
- multi class
- data sources
- prior knowledge
- training data
- text categorization
- data points
- probabilistic model
- cost function
- multi objective
- spectral clustering
- method for feature selection
- dimensionality reduction
- knowledge discovery
- probability distribution
- pairwise
- feature space
- machine learning