United States Air Force Officer Manpower Planning Problem via Approximate Dynamic Programming
Technical Report,01 Aug 2014,24 Mar 2016
AIR FORCE INSTITUTE OF TECHNOLOGY WRIGHT-PATTERSON AFB OH WRIGHT-PATTERSON AFB United States
Pagination or Media Count:
The United States Air Force USAF is concerned with managing its officer corps to ensure sufficient personnel for mission readiness. Manpower planning for the USAF is a complex process which requires making decisions about accessions, with the uncertainty about factors which could affect employee turnover. We formulate a Markov decision process model of the Air Force officer manpower planning problem AFO-MPP. We utilize the least squares approximate policy iteration algorithm as an approximate dynamic programming ADP technique to solve this large-scale problem. We test the performance of the policy determined with the ADP algorithm compared to a benchmark policy on two problem instances of the AFO-MPP. We find that the algorithm performs well for the basis functions selected, providing policies which reduce soft costs associated with shortages and surpluses in the force.
- Operations Research
- Logistics, Military Facilities and Supplies
- Personnel Management and Labor Relations