Wild Patterns Reloaded: A Survey of Machine Learning Security against Training Data Poisoning.
Antonio Emanuele CinàKathrin GrosseAmbra DemontisSebastiano VasconWerner ZellingerBernhard Alois MoserAlina OpreaBattista BiggioMarcello PelilloFabio RoliPublished in: CoRR (2022)
Keyphrases
- machine learning
- training data
- decision trees
- learning algorithm
- supervised learning
- training patterns
- support vector machine
- data sets
- data mining
- pattern recognition
- information security
- machine learning methods
- test data
- classification accuracy
- security issues
- artificial intelligence
- text classification
- pattern discovery
- semi supervised learning
- labeled data
- access control
- training set
- training process
- security requirements
- security policies
- knowledge discovery
- domain knowledge
- inductive learning
- training samples
- active learning
- sequential patterns
- explanation based learning
- cloud computing
- noisy data
- statistical databases
- learning problems
- learning tasks
- pattern mining
- privacy preserving
- training examples
- intrusion detection
- data mining techniques
- information extraction
- prior knowledge
- data analysis
- reinforcement learning
- feature selection
- neural network