Mathematical-Statistical and Digital Computer Analysis of Time Series Data.
Abstract
The purpose of this study is to develop a series of computer programs for use in analyzing time series data (waves) - such as EEG readings. Programs were used to produce reliable estimates of correlations, spectra, cross spectra, and partial coherences of multi-channel random processes. The software package was written to be easily adaptable to different sampling rates, amounts of data, and numbers of channels. Provisions for digital prefiltering of data, detrending, and smoothing (using a number of lag windows) were also included. Techniques for estimation of spectra by fitting single- and multi-channel autoregressive schemes to sampled data were also applied and found to yield results consistent with the other methods. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- May 17, 1976
- Accession Number
- ADA032719
Entities
People
- Steven J. Raher
Organizations
- United States Naval Academy