Exploratory Analysis of Economic Factors in the Navy Total Force Strength Model (NTFSM).

摘要

Accurate forecasts of U.S. Navy enlisted end-strength are crucial for budgetary planning and the development of manpower policies. An improving economy and increased employment opportunities in the civilian sector could cause a significant problem for enlisted retention. The Navy Total Force Strength Model (NTFSM) is a new stochastic simulation that is intended to offer manpower analysts more accurate enlisted manpower projections than those projected with the current tool. NTFSM uses historical data and user-defined inputs for economic factors to project monthly retention losses. However, NTFSM is still in the testing phase and its overall behavior is largely unknown. In particular, the analysts that NTFSM was designed to help are unsure of the effects that the economic factors, which they need to enter themselves, have on NTFSM's output. This thesis investigates the behavior of NTFSM's output and the sensitivity of the user-entered economic factors. Using design of experiments and data mining, a variety of scenarios are simulated and then analyzed to better understand the behavior of the model and to determine the sensitivity of the user-defined economic factors. The results of the analysis unexpectedly show that NTFSM s economic factors have no significant impact on NTFSM's end-strength output; this warrants further investigation.