The multiple criteria location problem : 1 . A generalized network model and the set of efficient solutions

This is the first of two papers in which multiple criteria location problems (MCLPs) are discussed. In this paper the main aim is to formalize a discrete MCLP and to develop a generalized network model. A critical overview of various techniques for generating efficient solutions to multiple criteria decision problems is offered. The three most commonly used methods for tackling MCLPs, namely the weighted method, the noninferior set estimation method, and the constraint method, are discussed. The main purpose of the generating techniques is to determine an exact representation of or an approximation to the set of efficient solutions among which one can choose the best or most preferred solution (location plan). To identify the best solution some information about the decisionmaker's preferences or a decision rule is needed. Consequently, in paper 2 we focus on preference-based approaches to multiple criteria decisionmaking and relate them to the concept of interactive decision support. Specifically, optimizing decision rules (utility-function-based approaches) and satisficing decision rules (goal programming methods) are discussed. Advantages and disadvantages of these two approaches to solving the MCLP are highlighted. It is suggested that the utility-maximizing and satisficing decision rules are not mutually exclusive. Accordingly, a quasi-satisficing approach that merges these two decision rules is proposed. Also, a framework for an interactive decision support system (DSS) for tackling MCLPs is presented. The system incorporates the generalized network model into a quasi-satisficing approach. It is argued that the DSS data and analytical components can be effectively integrated by means of the interactive decision support concept which allows for exploring the problem and the alternative solutions both in decision space and in criterion outcome space. 1 Introduction Locational decisionmaking can be considered as a process of searching for the best location or pattern of locations. This process can be divided into three stages: intelligence, design, and choice (Simon, 1960). The aim in the first phase is to identify the problem environment. It requires searching the environment for conditions which call for locational decisions. For example, an analysis of the geographical distribution of public service facilities in relation to the projected distribution of population may suggest the need for a spatial shift in the supply of public services and, consequently, for locational decisions. This type of activity requires an analysis of comprehensive spatially referenced data (Armstrong et al, 1992). To this end, the most valuable support for locational decisionmaking is offered by a …