Does the Order of Training Samples Matter? Improving Neural Data-to-Text Generation with Curriculum Learning.
Ernie ChangHui-Syuan YehVera DembergPublished in: CoRR (2021)
Keyphrases
- training samples
- learning algorithm
- supervised learning
- training data
- database
- data sets
- reinforcement learning
- feature space
- high quality
- background knowledge
- data points
- number of training samples
- test sample
- hyperplane
- data distribution
- learning process
- computer vision
- small number
- decision trees
- sample set
- artificial intelligence
- text generation