Summaries - Research
Back Improved Forecasting Methods for Naval Manpower Studies
|Division||Research & Sponsored Programs|
|Department||Naval Research Program|
|Investigator(s)||Koyak, Robert A.|
|Sponsor||NPS Naval Research Program (Navy)|
We propose a collaborative effort with NPRST to evaluate new statistical models for forecasting continuation, retention, and attrition rates. This effort will be shared so that the NPS investigators will develop two models and NPRST will develop one model. All models to be considered will be decided in collaboration with NPRST at the initial stage of the project. In addition, the NPS investigators will conduct a validation study of the model developed by NPRST using representative community data (i.e., aviation, submarine, and surface warfare communities).
During the development phase we will conduct an investigation into the statistical properties of two identified forecasting methods, using data and other information provided by NPRST. Our investigation will be guided by a set of metrics for evaluating and comparing the accuracy of forecasting methods that match the requirements of the Navy, recognizing that it is possible for a method to outperform another with respect to one metric, with the reverse being true for a different metric.
For validation of the forecasting method developed by NPRST we will use the identified metrics to characterize its retrospective performance on recent data, and we will use simulation (e.g. bootstrapping) where useful. The objective of this effort will be to compare the performance of the three forecasting methods under a common set of conditions so that their relative strengths and weakness can be identified.
|Publications||Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal|
|Data||Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal|