1984: AI - Next generation reservoir computing

[grarpamp]

Half the AI race is to analyze the world,
the other half, and ultimately the whole,
is to analyze and control you as part of it.

https://phys.org/news/2021-09-scientists-reservoir.html
https://www.nature.com/articles/s41467-021-25801-2
https://doi.org/10.1038/s41467-021-25801-2
https://github.com/quantinfo/ng-rc-paper-code

Next generation reservoir computing
Gauthier, D.J.

Reservoir computing is a best-in-class machine learning algorithm for
processing information generated by dynamical systems using observed
time-series data. Importantly, it requires very small training data
sets, uses linear optimization, and thus requires minimal computing
resources. However, the algorithm uses randomly sampled matrices to
define the underlying recurrent neural network and has a multitude of
metaparameters that must be optimized. Recent results demonstrate the
equivalence of reservoir computing to nonlinear vector autoregression,
which requires no random matrices, fewer metaparameters, and provides
interpretable results. Here, we demonstrate that nonlinear vector
autoregression excels at reservoir computing benchmark tasks and
requires even shorter training data sets and training time, heralding
the next generation of reservoir computing.