DeepAstroUDA: semi-supervised universal domain adaptation for cross-survey galaxy morphology classification and anomaly detection.
Aleksandra CiprijanovicAshia LewisKevin PedroSandeep MadireddyBrian NordGabriel N. PerdueStefan M. WildPublished in: Mach. Learn. Sci. Technol. (2023)
Keyphrases
- anomaly detection
- domain adaptation
- semi supervised
- unsupervised anomaly detection
- unsupervised learning
- labeled data
- co training
- semi supervised learning
- supervised learning
- intrusion detection
- unlabeled data
- anomalous behavior
- class labels
- document classification
- active learning
- detecting anomalies
- network intrusion detection
- text classification
- intrusion detection system
- multiple sources
- label information
- decision trees
- machine learning
- one class support vector machines
- support vector machine
- feature vectors
- pairwise
- support vector
- neural network
- multi view
- transfer learning
- detect anomalies
- training set
- pattern recognition
- classification algorithm
- learning algorithm
- data analysis
- target domain
- feature representation
- cross domain
- prior knowledge