Defending SVMs against poisoning attacks: the hardness and DBSCAN approach.
Hu DingFan YangJiawei HuangPublished in: UAI (2021)
Keyphrases
- ddos attacks
- support vector
- density based clustering
- clustering algorithm
- countermeasures
- clustering method
- multi class
- data clustering
- classification using support vector machines
- kernel function
- feature selection
- support vector machine svm
- structured learning
- k means
- training support vector machines
- maximum margin
- machine learning
- support vector machine
- sparse kernel
- security threats
- feature space
- density based clustering algorithm
- training data
- multi class classification
- computational complexity
- watermarking technique
- information theoretic
- np complete
- support vectors
- generalization ability
- hyperplane
- learning theory
- phase transition
- cost sensitive
- svm classifier
- security mechanisms
- training set
- feature vectors
- decision trees
- spatial clustering
- reduced set
- logistic regression
- kernel methods