Low-Power Anomaly Detection and Classification System based on a Partially Binarized Autoencoder for In-Sensor Computing.
Paola VitoloGian Domenico LicciardoLuigi Di BenedettoRosalba LiguoriAlfredo RubinoDanilo PauPublished in: ICECS (2021)
Keyphrases
- anomaly detection
- low power
- high speed
- low cost
- power consumption
- intrusion detection
- unsupervised learning
- image sensor
- anomalous behavior
- network intrusion detection
- network traffic
- single chip
- detecting anomalies
- network anomaly detection
- decision trees
- feature vectors
- pattern recognition
- text classification
- machine learning
- support vector
- intrusion detection system
- vlsi circuits
- model selection
- feature selection
- network security
- sensor networks
- feature extraction
- one class support vector machines
- logic circuits
- maximum entropy
- negative selection algorithm
- detecting anomalous
- detect anomalies
- low power consumption
- power reduction
- neural network
- misuse detection
- power dissipation
- supervised learning
- support vector machine
- object recognition
- data mining