Moving Average Models--Time Series in M-Dimensions.
Abstract
Stochastic models for discrete time series in the time domain are well known but such models lack consideration of spatial dependency. We expand on their work by constructing spatially depending moving average models. Definitions of order, stationarity, invertibility, autocorrelation function, and spectrum are made as natural extensions of those in zero dimensions and are implemented in the one and two-space dimensional models. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 15, 1978
- Accession Number
- ADA054585
Entities
People
- C. A. Oprian
- D. A. Voss
- L. A. Aroian
Organizations
- Union College