Non-linear prediction with LSTM recurrent neural networks for acoustic novelty detection.
Erik MarchiFabio VesperiniFelix WeningerFlorian EybenStefano SquartiniBjörn W. SchullerPublished in: IJCNN (2015)
Keyphrases
- recurrent neural networks
- novelty detection
- long short term memory
- chaotic time series
- text filtering
- neural network
- feed forward
- anomaly detection
- echo state networks
- recurrent networks
- reservoir computing
- artificial neural networks
- feedforward neural networks
- sentence level
- artificial intelligence
- support vector data description
- cascade correlation
- bayesian networks
- genetic algorithm
- concept drift
- remote sensing
- nonlinear dynamic systems
- text categorization
- data structure