Active Learning for the Optimal Design of Multinomial Classification in Physics.
Yongcheng DingJosé D. Martín-GuerreroYujing SongJosé Rafael Magdalena BeneditoXi ChenPublished in: CoRR (2021)
Keyphrases
- optimal design
- active learning
- text classification
- selective sampling
- pattern classification
- classification accuracy
- batch mode active learning
- classification algorithm
- decision trees
- artificial intelligence
- water supply
- classification scheme
- image classification
- machine learning
- probabilistic model
- feature extraction
- training set
- machine learning methods
- structural design
- support vector
- experimental design
- feature space
- classification method
- feature vectors
- naive bayes
- labeled data
- support vector machine
- supervised learning
- learning strategies
- transfer learning
- semi supervised
- cost sensitive
- classification models
- random sampling
- training samples
- unlabeled data
- unsupervised learning
- feature selection
- annotation effort