Non- and semiparametric methods for surface estimation of functional time series on a lattice / von Bastian Schäfer M. Sc. Paderborn, 2022
Content
- Contents
- List of Figures
- List of Tables
- List of Abbreviations and Acronyms
- Introduction
- Spatial Time Series and Nonparametric Regression
- Contribution of this Thesis
- Summary of the Contents
- Boundary Modification in Local Regression
- Introduction
- Modified Local Polynomial Regression
- Equivalency of the Proposed Weighting Methods
- Boundary Behavior of Local Regression
- Areas of Application
- Final Remarks
- Appendix
- Fast Computation and Bandwidth Selection Algorithms for the DCS
- Introduction
- The Model and the Basic DCS Procedure
- The Improved Double Conditional Smoothing
- Boundary Correction Under the DCS
- A Functional Smoothing Scheme
- Estimation of Derivatives
- Asymptotic Behavior of the Estimator
- Bandwidth Selection
- Finite Sample Simulations
- Application to Financial Data
- Final Remarks
- Appendix
- Local Polynomial DCS under Dependent Errors
- Introduction
- The FDCS for Local Polynomial Estimators
- Model and Assumptions
- Extension to Local Polynomial Smoothers
- Boundary Modification in the LP-DCS
- Equivalent Kernels
- Bandwidth Selection for the LP-DCS
- Spatial Error Structure
- Simulation Study
- Applications
- Final Remarks
- Appendix
- Further Research Topics
- Conclusion
- Appendix: DCSmooth Vignette
- Introduction
- Details of Functions, Methods and Data
- Application
- Defining the Options
- Application of the DCS with iid. Errors
- Application of the DCS with SARMA Errors
- Modeling Errors with Long Memory
- Estimation of Derivatives
- Mathematical Background
- Bibliography