THEORY AND APPLICATION OF AN INFERENCE MODEL FOR NON-STATIONARY TIME SERIES MEANS.
Abstract
Complex time series models, such as consumer brand shifting analyses, have required assumptions of parameter stability because statistical models to deal with parameter change were not available. A model is developed here to make inferences about a possibly nonstationary time series mean generating data with Gaussian error. Estimators which are efficient in a special sense are presented, along with examples and suggested applications of the method to brand switching problems. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1964
- Accession Number
- AD0613369
Entities
People
- John U. Farley
- Melvin Hinich
Organizations
- Carnegie Institute of Technology