Development of an accelerating hungarian method for assignment problems

Abstract

The Hungarian method is a well-known method for solving the assignment problem. This method was developed and published in 1955. It was named the Hungarian method because two theorems from two Hungarian mathematicians were used. In 1957, it was noticed that this algorithm is strongly polynomial and has a complexity of order O(n 4 ) This is the reason why the Hungarian method is also known as the Kuhn-Munkres algorithm. Later on, in 1971 the complexity of the method was improved to order O(n 3 ) A smallest uncovered element is selected to create a single zero at every iteration. This is a weakness and is alleviated by selecting more than one smallest uncovered element thus creating more than one zero at every iteration to come up with what we now call the Accelerating Hungarian (AH) method. From the numerical illustration of the Hungarian method given in this paper, we require 6 iterations to reach optimality. It can also be shown that selecting a single smallest uncovered element (e s ) makes the Hungarian method inefficient when creating zeros. From the same numerical illustration of the proposed algorithm (AH) also given in this paper, it can be noted that only one iteration is required to reach optimality and that a total of six zeros are created in one iteration. Assignment model and the Hungarian method have application in addressing the Weapon Target Assignment (WTA) problem. This is the problem of assigning weapons to targets while considering the maximum probability of kill. The assignment problem is also used in the scheduling problem of physicians and medical staff in the outpatient department of large hospitals with multi-branches. The mathematical modelling of this assignment problem results in complex problems. A hybrid meta-heuristic algorithm SCA–VNS combining a Sine Cosine Algorithm (SCA) and Variable Neighbourhood Search (VNS) based on the Iterated Hungarian algorithm is normally used.