Title page for ETD etd-05112005-123417


Type of Document Master's Thesis
Author Ames, Allison Jennifer
Author's Email Address aames@vt.edu
URN etd-05112005-123417
Title Monte Carlo Experiments on Maximum entropy Constructive Ensembles for Time Series Analysis and Inference
Degree Master of Science
Department Agricultural and Applied Economics
Advisory Committee
Advisor Name Title
Hilmer, Christiana E. Committee Chair
Spanos, Aris Committee Co-Chair
Taylor, Daniel B. Committee Member
Keywords
  • maximum entropy
  • ensembles
  • time series
  • bootstrap
Date of Defense 2005-05-09
Availability unrestricted
Abstract
In econometric analysis, the traditional bootstrap and related methods often require the assumption of stationarity. This assumption says that the distribution function of the process remains unchanged when shifted in time by an arbitrary value, imposing perfect time-homogeneity. In terms of the joint distribution, stationarity implies that the date of the first time index is not relevant. There are many problems with this assumption however for time series data. With time series, the order in which random realizations occur is crucial. This is why theorists work with stochastic processes, with two implicit arguments, w and t, where w represents the sample space and t represents the order. The question becomes, is there a bootstrap procedure that can preserve the ordering without assuming stationarity? The new method for maximum entropy ensembles proposed by Dr. H. D. Vinod might satisfy the Ergodic and Kolmogorov theorems, without assuming stationarity.

Files
  Filename       Size       Approximate Download Time (Hours:Minutes:Seconds) 
 
 28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access 
  MonteCarloExperimentsonMEConstructiveEnsemblesforTimeSeriesAnalysisandInference.pdf 355.44 Kb 00:01:38 00:00:50 00:00:44 00:00:22 00:00:01

Browse All Available ETDs by ( Author | Department )

dla home
etds imagebase journals news ereserve special collections
virgnia tech home contact dla university libraries

If you have questions or technical problems, please Contact DLA.