Time Series Models for Birth Forecasting.
Abstract
Autoregressive integrated moving average (ARIMA) models are developed for the birth time series, and their relationship with the classical models for population growth is investigated. Parsimonious versions of the ARIMA models are obtained which retain the most important pieces of information including the length of generation of the population. The technique is applied to human population data (Mexico and Norway) and forecasts are obtained. A causal model relating marriages to births is also developed and applied.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1975
- Accession Number
- ADA010145
Entities
People
- Joao L. M. Saboia
Organizations
- University of California, Berkeley