Robust learning of data anomalies with analytically-solvable entropic outlier sparsification.
Illia HorenkoPublished in: CoRR (2021)
Keyphrases
- data sets
- data collection
- knowledge discovery
- prior knowledge
- learning process
- high quality
- statistical analysis
- data analysis
- raw data
- data structure
- learning algorithm
- data quality
- database
- synthetic data
- anomaly detection
- neural network
- small number
- data points
- data sources
- training data
- bayesian methods
- previously learned
- outlier detection
- sensor data
- background knowledge
- original data
- hidden variables
- detection algorithm
- data processing
- data mining techniques
- supervised learning
- probability distribution
- end users
- special case
- data mining
- computational complexity
- learning environment