Multiple sources of noise can affect the encoding of information in the nervous
system. We assessed the robustness of the code to neuronal drop-out. The classifier
was trained with all sites (here n=256) as described in the Methods.
Here, before testing the performance of the classifier, a proportion of sites
(indicated in the x-axis) were removed from the classifier (simulating the process
of neuronal death or axonal death). Classifier parameters: MUA, n=256 sites,
[100;300) ms time interval, bin size = 50 ms.