Data pruning and neural scaling laws: fundamental limitations of score-based algorithms.
Fadhel AyedSoufiane HayouPublished in: CoRR (2023)
Keyphrases
- data processing
- data collection
- optimization problems
- data structure
- image data
- database
- data mining algorithms
- synthetic data
- data analysis
- data sets
- data mining techniques
- learning algorithm
- data reduction
- incomplete data
- computational complexity
- high quality
- training data
- computer systems
- statistical analysis
- neural network
- experimental data
- noisy data
- data mining
- discrete data
- data quality
- original data
- raw data
- data sources
- nearest neighbor
- prior knowledge
- probability distribution