Statistical models for spike trains:
some thoughts on the current situation
Department of Statistics and
Center for the Neural Basis of Cognition, Carnegie Mellon University
I will discuss some of the problems involved
in analyzing spike train data, together with LFPs,
from substantial numbers of simultaneously-recorded
neurons. For off-line analysis, the main topic I'll focus on
is description of correlated activity. I will briefly review
available statistical modeling approaches,
and the scientific issues they aim to deal with,
and will try to identify currently unmet needs.
I also expect to say a few words about on-line analysis
for neural prostheses.