Figure 7.

Comparison between different probability density functions (PDFs) using higher order moments and the local variability (LV). Higher order moments of the empirical distributions and the local variability of the spike trains can be employed to validate the use of specific PDFs as approximations of the empirical distributions as they inform on features that cannot be seen using only the mean and standard deviation. (A,B) display the attained 99% confidence bounds of the constructed model using the gamma, lognormal and inverse Gaussian distribution. (A) Spinal interneuron that followed the predicted LV of the gamma distribution. (B) Representative spinal interneuron with no preferred PDF. (C) Summary of all tested neurons showing the proportion of stationary regions that fall within the confidence bounds of the respective model. The slight trend in (B) where the gamma distribution underestimates the L-skewness and L-kurtosis is more evident here, while there is no difference between the PDFs regarding the LV.