Robust Reservoir Generation by Correlation-Based Learning
Figure 2
The basic property of the
reservoir in our RC model. (a) Active states of the first 100 out of excitatory neurons (. The abscissa and ordinate represent time
and neuron index, respectively. A horizontal line represents the duration of
the active state of a neuron. (b) The similarity index of this activity
pattern. The abscissa and ordinate are time difference () and the index, respectively. The solid line
indicates the mean index value, whereas the gray region represents the standard
deviation of the index. (c) The reproducibility index between two sequences of
active neuron populations generated under different noise. The abscissa and
ordinate represent time and the index, respectively.