The expressive power of analog recurrent neural networks on infinite input streams

@article{Cabessa2012TheEP,
  title={The expressive power of analog recurrent neural networks on infinite input streams},
  author={J{\'e}r{\'e}mie Cabessa and Alessandro E. P. Villa},
  journal={Theor. Comput. Sci.},
  year={2012},
  volume={436},
  pages={23-34}
}
We consider analog recurrent neural networks working on infinite input streams, provide a complete topological characterization of their expressive power, and compare it to the expressive power of classical infinite word reading abstract machines. More precisely, we consider analog recurrent neural networks as language recognizers over the Cantor space, and… CONTINUE READING