Pareto Frontiers in Deep Feature Learning: Data, Compute, Width, and Luck.
Benjamin L. EdelmanSurbhi GoelSham M. KakadeEran MalachCyril ZhangPublished in: NeurIPS (2023)
Keyphrases
- raw data
- prior knowledge
- database
- data analysis
- high dimensional data
- active learning
- data sources
- data sets
- data distribution
- synthetic data
- statistical analysis
- data collection
- supervised learning
- learning algorithm
- data structure
- probability distribution
- deep learning
- machine learning
- reinforcement learning
- computer systems
- data processing
- learning process
- image data
- high quality
- decision trees
- missing data
- background knowledge
- missing values
- human experts
- knowledge discovery
- noisy data
- data quality
- learning models
- hidden variables
- multi objective