Comparison of Echo State Networks with Simple Recurrent Networks and Variable-Length Markov Models on Symbolic Sequences.
Michal CernanskýPeter TiñoPublished in: ICANN (1) (2007)
Keyphrases
- variable length
- markov models
- echo state networks
- recurrent neural networks
- recurrent networks
- fixed length
- hidden markov models
- markov model
- sequence classification
- higher order
- n gram
- maximum entropy
- neural network
- bitstream
- artificial neural networks
- conditional random fields
- feed forward
- fuzzy logic
- similarity measure
- feature extraction