A multi-objective framework for the project selection problem using goal programming

Abstract

It has been pointed out in the literature that usually there are multiple objectives relevant to the problem of selecting a flow project from many project proposals. Barring some recent work in this field, the project selection problem has generally been treated as one having a single objective in formal decision models. Goal Programming, with its satisficing as opposed to the maximizing approach , offers a relatively simple way of handling such problems; however, the method of pre-amptive weight which has generally been used to solve this problem makes a severely restrictive assumption about the decision maker having an ordinal goal hierarchy and pre-cludes any effective trade-off between goals. In this study the project selection problem as faced by a public financial institution has been analyzed and formulated as a 0-1 minmax Goal Programming problem. Some theoretical properties of the minmax Goal Programming formulation have been studied and solution procedures developed for solving the same. The applicability of the formulation and the solution procedures has been demonstrated with real life data on projects taken from a public financial institution. Four goals, viz. budget goal, npv goal, employment goal and investment in backward area goal, have been used in the demonstration. The main contribution of this research, in the areas of theory and applications of Goal programming are as a follows; i) Extending the concept of efficiency and efficient sets to the area of Goal Programming and developing a theoretical basis for the minmax Goal Programming formulation, Which minimizes of the many weighted goal deviations. ii) Developing a solution procedure for the 0-1 minmax Goal Programming problem based on a modification of the Balas’ implicit enumeration procedure. iii) Developing a solution procedure for the minmax linear Goal Programming which uses the special structure of the formulation and enables us to solve the problem while actually solving a much reduced problem. iv) Providing a multi- objective framework for the project selection problem as faced by a public financial institution.