Corpus ID: 219260808

dynoNet: a neural network architecture for learning dynamical systems

@article{Forgione2020dynoNetAN,
  title={dynoNet: a neural network architecture for learning dynamical systems},
  author={Marco Forgione and D. Piga},
  journal={ArXiv},
  year={2020},
  volume={abs/2006.02250}
}
  • Marco Forgione, D. Piga
  • Published 2020
  • Computer Science, Engineering, Mathematics
  • ArXiv
  • This paper introduces a network architecture, called dynoNet, utilizing linear dynamical operators as elementary building blocks. Owing to the dynamical nature of these blocks, dynoNet networks are tailored for sequence modeling and system identification purposes. The back-propagation behavior of the linear dynamical operator with respect to both its parameters and its input sequence is defined. This enables end-to-end training of structured networks containing linear dynamical operators and… CONTINUE READING

    Figures and Topics from this paper.

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 23 REFERENCES
    Time series classification from scratch with deep neural networks: A strong baseline
    304
    LSTM: A Search Space Odyssey
    1969
    Adam: A Method for Stochastic Optimization
    47438
    Generalised Hammerstein–Wiener system estimation and a benchmark application
    40
    Parametric identification of parallel Wiener-Hammerstein systems
    27
    Automatic differentiation in machine learning: a survey
    444