Characterizing Spontaneous Neurophysiological Activity with Measures of Statistical Serial Dependence: a Summary Statistic and an Extension of the Joint Interval Histogram

Autor: Weil, Jon C. (Jon Christopher)
Jazyk: angličtina
Rok vydání: 1994
Předmět:
Druh dokumentu: Thesis or Dissertation
Popis: Two measures which indicate statistical serial dependence were evaluated. The n-dimensional Average Stored Information index (ndASI) is a measure of conditional information, which compares the entropies of higher order conditional distributions to estimate average statistical serial dependence. The generalized ranked joint interval histogram (RJIH-o) is a new nonparametric graphical analysis method. It extends the joint interval histogram by depicting longer interval sequences and can be interpereted precisely analagously to the joint interval histogram. The generalized ranked joint interval histogram correctly represents independence in a Poisson process model and statistical serial dependence in a directionally reinforced Markov model. The generalized ranked joint interval histogram correctly depicts the underlying periodic and strange attractors in a standard map model. Both measures can be used to effectively analyze interspike interval sequences from spontaneous neurophsyiological activity effectively.
Databáze: Networked Digital Library of Theses & Dissertations