On modeling dynamic priorities in the analytic hierarchy process using compositional data analysis
The subject of time-dependent priorities is a new and potentially useful development in the AHP/ANP literature. It involves making paired comparisons not only of relative dominance of magnitudes as a starting point, but also of rates of change. Dealing with the future of human interaction and decisionmaking is very uncertain and confounding subject. We need caution in developing the subject if it is to work out to our advantage. This paper is mainly concerned with the theory of the subject. Examples are given to illustrate the method and detailed discussion is provided about how to derive time dependent priorities analytically for matrices of order up to four and numerically for matrices of all order. A brief representation and synthesis of time dependent neural firing and its hypermatrix is given.