Sankhya: The Indian Journal of Statistics

1995, Volume 57, Series B, Pt. 1, pp. 32--47



SHUBHABRATA DAS, University of Montana, Missoula


PRANAB KUMAR SEN, University of North Carolina, Chapel Hill

SUMMARY.  The Central Nervous System has different sectors controlling different activities, and it is possible to observe simultaneously spike-trains for multiple nerve calls. Stochastic dependence of spike-trains within a small sector may differ from that between two such sectors. Canonical correlation analysis plays a vital role in this study. In a nonparametric setup, incorporating binning, suitable canonical dependence measures are proposed, and their asymptotic distributions are studied. Some wellknown resampling plans are incorporated to estimate the bias and variance.

AMS (1990) subject classification.  60G35, 62H20, 62P99.

Key words and phrases. Asymptotics, binning, bootstrap, canonical correlations, continuous time parameter, counting process, intensity function, jackknife, neural network, neurophysiology, resampling plans, restricted canonical correlations.

Full paper (PDF)

This article in mathematical reviews.