Effective prediction of intercensal population levels

Abstract

This dissertation is concerned with the prediction of population levels for the years following a census until the next census is counted. It reviews standard interpolation and extrapolation tecnhiques, population models and neural networks. The population levels are required by government for allocating money to local authorities for spending on local services. This project was initiated by Chester City Council who consider that an underestimation of the Chester levels is causing a shortfall in the city's allocation of money.