A Dimensionality-Driven Approach for Unsupervised Out-of-distribution Detection.
Qizhou WangSarah M. ErfaniChristopher LeckieMichael E. HoulePublished in: SDM (2021)
Keyphrases
- data driven
- detection method
- automatic detection
- unsupervised learning
- detection rate
- dimensionality reduction
- detection accuracy
- false alarms
- detection algorithm
- novelty detection
- false positives
- semi supervised
- pairwise
- supervised learning
- probability distribution
- high dimensional
- data distribution
- high dimensionality
- feature space
- decision trees
- neural network
- data sets
- spatial distribution
- small size
- real time