Data-splitting improves statistical performance in overparameterized regimes.
Nicole MückeEnrico ReissJonas RungenhagenMarkus KleinPublished in: AISTATS (2022)
Keyphrases
- data sets
- database
- high quality
- data collection
- data sources
- image data
- raw data
- application domains
- statistical analysis
- experimental data
- complex data
- data distribution
- missing values
- statistical methods
- synthetic data
- computer systems
- data processing
- original data
- data analysis
- training data
- genetic algorithm
- data objects
- statistical information
- statistical data
- spatial data
- missing data
- privacy preserving
- data mining techniques
- semi supervised
- knowledge discovery
- data structure
- real time