A decision support system for resource allocation derived from Poisson gravity regression analysis and linear goal programming

摘要

The authors demonstrate the application of a methodology derived from Poisson gravity regression and linear goal programming in allocating funds for leases among different shopping malls in the Atlanta retailing system. It is indicated that a retail store's profitability is directly related to the volume of sales generated and sales volume in turn is determined by patronization rate, income of shoppers, and the drawing power of anchor tenants. The Poisson gravity model was initially used to forecast the patronization rates for different malls. Using these forecasts as the objectives the decision maker wants to achieve, a linear goal programming (LGP) model was constructed. The results of the LGP indicate that combining a method to forecast patronization rates with a multiple objective resource allocation methodology such as LGP leads to a satisfactory distribution of lease funds. Specifically, a retail center with a relatively higher contribution to the objectives specified by the decision maker received relatively more funds.