Finding needles in a haystack: Sampling Structurally-diverse Training Sets from Synthetic Data for Compositional Generalization.
Inbar OrenJonathan HerzigJonathan BerantPublished in: CoRR (2021)
Keyphrases
- synthetic data
- training set
- real world
- data sets
- real image data
- mri data
- monte carlo
- wide variety
- classification accuracy
- neural network
- inductive bias
- feature space
- decision trees
- data mining
- multi class
- real life
- supervised learning
- training data
- random sampling
- classification error
- markov chain monte carlo
- sampling algorithm
- machine learning