Spelling suggestions: "subject:"egression."" "subject:"aregression.""
71 |
Latent structure models for repeated measurements experimentsWhite, S. A. January 1986 (has links)
No description available.
|
72 |
The parametrisation of statistical modelsHills, Susan January 1989 (has links)
No description available.
|
73 |
Regression using QR decomposition methodsSmith, David McCulloch January 1991 (has links)
No description available.
|
74 |
Methods of inference for nonparametric curves and surfacesBock, Mitchum T. January 1999 (has links)
No description available.
|
75 |
Bayesian inference for functionsUpsdell, M. P. January 1985 (has links)
No description available.
|
76 |
Measurement errors in case-control and related studiesGunby, James Alexander January 1993 (has links)
No description available.
|
77 |
Updating the Navy's recruit quality matrix an analysis of educational credentials and the success of first-term sailorsBownds, Christopher D. 03 1900 (has links)
Approved for public release; distribution is unlimited. / This thesis analyzes the impact of different
educational credentials on first-term attrition by enlisted
sailors in the U.S. Navy. For enlistment screening, the
Navy currently categorizes applicants in one of three tiers
according to educational attainment. These tiers form the
basis of the Recruit Quality Matrix, which employs Armed
Forces Qualification Test scores and educational
credentials to determine enlistment eligibility. The
analysis draws primarily from two sources: a Defense
Manpower Data Center file containing enlisted cohorts from
fiscal years 1989 through 1997 (to assess first-term
attrition), and a Commander, Navy Recruiting Command data
base containing enlisted cohorts from fiscal years 1998
through 2003 (to examine bootcamp attrition). Logit
regression models are constructed using these data to
identify differences in attrition propensities within the
general tiers. A refined matrix is designed and evaluated
as a more accurate predictor of attrition. Further research
is recommended to look at additional measures of success in
service, such as performance, productivity, and promotion. / Lieutenant Commander, United States Navy
|
78 |
Forecasting retention in the United States Marine Corps ReserveSchumacher, Joseph F. 03 1900 (has links)
This is an empirical study using a logistic regression model to assess the impact of mobilization and unemployment on an individual's decision to stay in or leave the reserves. The goal is to find out the attrition behavior of USMCR participants in order to better establish recruiting and retention goals in the Reserve population. Questions regarding attrition influencers, effects of mobilization, and applicability to both officer and enlisted personnel were reviewed in this process. The effects of being called to active service are shown to have a positive effect on retention in the reserves. Similarly, serving in the SMCR and Stand-by Reserves are both shown in the model to have a positive effect on reserve retention. This makes sense, in that when an individual volunteers in the Marine Reserves, he or she evidences a desire to serve his country when called to do so. The negative effect of an increase in the number of days served on active duty, as shown in the results of the model, follows similar logic. Had the individual wanted to serve on a full-time active duty basis he would have volunteered for the active duty component. The longer he is asked to remain on active duty, the more dissatisfied he is, on average, with his participation in the reserves. The negative effect of an increase in the individual's home of record unemployment rate is also consistent with previous findings, and when combined with the negative effect of continued mobilization and recall from the IRR or a retired status, a significant negative impact is seen on the individual's decision to stay in. The findings indicate that multiple short activations have a positive impact, whereas the impact of fewer, lengthy activations is negative This study validated previous research regarding the likelihood to continue to serve in the Marine Corps Reserves. As a result, the Marine Corps has the potential to better allocate resources and schedule individual activations, reducing attrition. This can assist in shaping the force structure when the Marine Corps are needed.
|
79 |
Nejmenší vážené čtverce a jejich asymptotika / The least weighted squares and its asymptoticsRaušová, Magdaléna January 2016 (has links)
When there are some influential observations present in a data set (such as outliers or leverage points), the use of some robust method may be desirable for being able to draw relevant conclusions from an econometric analysis. In order to use these methods properly, we need some diagnostic tools. To be able to derive these tools theoretically, we first need to know the form of the asymptotic representation of corresponding estimator. This thesis derives the asymptotic representation of the estimator obtained by the method of least weighted squares under the assumption of heteroskedastic residuals. The tight- ness of the estimator and its asymptotic representation under several levels of contamination is also shown in a simulation study.
|
80 |
A comparison study on the estimation in Tobit regression modelsLeiker, Antoinette January 1900 (has links)
Master of Science / Department of Statistics / Weixing Song / The goal of this report is to compare various estimation procedures on regression models in which the dependent variable has a restricted range. These models, called Tobit models, are seeing an increase in use among economists and market researchers, specifically. Only the standard Tobit regression model is discussed in the report.
First we will examine the five estimation methods discussed in Amemiya (1984) for standard Tobit model. These methods include Probit maximum likelihood, least squares, Heckman’s two-step, Tobit maximum likelihood, and the EM algorithm. We will examine the algorithm utilized in each method’s estimation process.
We will then conduct simulation studies using these estimation procedures. Twelve scenarios have been considered consisting of three different truncation threshold on the response variable, two distributions of
covariates, and the error variance known and unknown. The results are reported and a discussion of the goodness of each method follows.
The study shows that the best method for estimating Tobit regression models is indeed the Tobit maximum likelihood estimation. Heckman’s two-step method and the EM algorithm also estimate these models well when the truncation rate is low and the sample size is large. The simulation results show that the Least squares estimation procedure is far less efficient than other estimation procedures.
|
Page generated in 0.0878 seconds