Domain Adaptation Principal Component Analysis: base linear method for learning with out-of-distribution data.
Evgeny M. MirkesJonathan BacAziz FouchéSergey V. StasenkoAndrei Yu. ZinovyevAlexander N. GorbanPublished in: CoRR (2022)
Keyphrases
- domain adaptation
- test data
- prior knowledge
- data sets
- principal component analysis
- data sources
- data distribution
- training samples
- pairwise
- learning process
- knowledge discovery
- training and test data
- training data
- covariate shift
- cross domain
- multiple sources
- support vector machine
- error rate
- feature set
- data points
- feature space
- data analysis
- similarity measure