A Data Utility-Driven Benchmark for De-identification Methods.
Oleksandr TomashchukDimitri Van LanduytDaniel PleteaKim WuytsWouter JoosenPublished in: TrustBus (2019)
Keyphrases
- data sets
- knowledge discovery
- data analysis
- high dimensional data
- data processing
- missing values
- statistical methods
- training data
- high quality
- data points
- statistical tests
- data mining techniques
- data mining methods
- human experts
- experimental data
- noisy data
- original data
- data collection
- small number
- data structure
- data reduction
- neural network
- data quality
- historical data
- disclosure risk
- raw data
- benchmark datasets
- synthetic data
- image data
- significant improvement
- machine learning
- databases