Clustering Approach for Detecting Multiple Types of Adversarial Examples.
SeokHwan ChoiTae-u BahkSungyong AhnYoon-Ho ChoiPublished in: Sensors (2022)
Keyphrases
- multiple types
- heterogeneous information networks
- heterogeneous networks
- real world
- co occurrence
- contextual information
- unsupervised learning
- training examples
- data points
- class labels
- networked data
- multiple features
- data objects
- data fusion
- dimensionality reduction
- multi class
- user interface
- mobile devices
- support vector
- similarity measure