The Liquid State Machine (LSM) (Maass et. al. 2002) is a computational model which uses the high-dimensional, complex dynamics of recurrent neural circuits to conduct memory-dependent readout operations on continuous input streams.
This package provides a convenience wrapper for network construction, as well as typical operations on the reservoir.
Note that this code does not implement the exact model in (Maass et. al. 2002), but rather a slightly simplified model. This code was used in (Kaiser et al. 2017).
After cloning this repository locally, run pip install . in the working copy. Requires a working installation of the NEST simulator.