| StempCens-package {StempCens} | R Documentation |
Spatio-Temporal Estimation and Prediction for Censored/Missing Responses
Description
It estimates the parameters of a censored or missing data in spatio-temporal models using the SAEM algorithm (Delyon et al., 1999). This algorithm is a stochastic approximation of the widely used EM algorithm and an important tool for models in which the E-step does not have an analytic form. Besides the expressions obtained to estimate the parameters to the proposed model, we include the calculations for the observed information matrix using the method developed by Louis (1982). To examine the performance of the fitted model, case-deletion measure are provided.
Details
The functions provided are:
- CovarianceM: computes the spatio-temporal covariance matrix for balanced data.
- EffectiveRange: computes the effective range for an isotropic spatial correlation function.
- EstStempCens: returns the maximum likelihood estimates of the unknown parameters.
- PredStempCens: performs spatio-temporal prediction in a set of new S spatial locations for fixed time points.
- CrossStempCens: performs cross-validation, which measure the performance of the predictive model on new test dataset.
- DiagStempCens: returns measures and graphics for diagnostic analysis.
Author(s)
Larissa A. Matos (ORCID), Katherine L. Valeriano (ORCID) and Victor H. Lachos (ORCID)
Maintainer: Larissa A. Matos (larissa.amatos@gmail.com).
References
Cook R (1977). “Detection of influential observation in linear regression.” Technometrics, 19(1), 15–18. doi: 10.1080/00401706.1977.10489493.
Delyon B, Lavielle M, Moulines E (1999). “Convergence of a stochastic approximation version of the EM algorithm.” Annals of Statistics, 27(1), 94–128. doi: 10.1214/aos/1018031103.
Louis T (1982). “Finding the observed information matrix when using the EM algorithm.” Journal of the Royal Statistical Society: Series B (Methodological), 44(2), 226–233. doi: 10.1111/j.2517-6161.1982.tb01203.x.
Zhu H, Lee S, Wei B, Zhou J (2001). “Case-deletion measures for models with incomplete data.” Biometrika, 88(3), 727–737. doi: 10.1093/biomet/88.3.727.