Optimizing Human Resource Management Models

We propose the study of a new heuristic for the general mixed integer problem called the Neighborhood Covering Heuristic (NCH). We describe the heuristic in its simplest, worst-case form. We then describe three enhancements to the heuristic which,separately and in combination, should yield a practical method for finding solutions. We believe this approach has the potential to be significantly faster and more effective than current offerings in finding solutions in very large problem spaces. ThisPhase I project will propose a method for demonstrating whether the approach is feasible and whether it is superior to existing approaches such as the Branch and Bound (BAB) method. Many real-world decisions can be cast as mixed integer programming (MIP)problems, including resource allocation, blending, and scheduling. Commonly known forms of the general MIP problem are the knapsack problem, the assignment problem, the lockbox problem, the network flow problem, and the set-covering problem. Theapplication areas for techniques for solving the MIP problem include human resource planning, financial planning, manufacturing, distribution, inventory, airline scheduling, telecommunications network planning, and many others. All of these disciplines inboth the military and commercial workplace require computer software to aid in solving these complex but common business challenges. This STTR project will, if successful, provide a software package using the proposed Neighborhood Covering Heuristic thatis significantly faster and more effective than current methods in finding solutions to these important large scale business problems.