Year of Publication
Doctor of Philosophy (PhD)
Arts and Sciences
Dr. Derek S. Young
Confidence intervals are used to capture a parameter of interest, usually a mean or a quantile, at a specified confidence level. Prediction intervals are another practical interval that aim at making sound predictions of future values with some confidence. Although these are useful inference tools, neither of them gives people a plausible range of the sampled population. Tolerance intervals are such an inference tool that captures a specified proportion of the sampled population at a predetermined confidence level. In this dissertation, tolerance intervals for an autoregressive process with order p are constructed. In addition, a method of utilizing tolerance interval to find out appropriate cutoff values is also comprehensively investigated.
Digital Object Identifier (DOI)
Cheng, Kedai, "Tolerance Intervals for Time Series Models and Specifying Trimming/Winsorizing Cutoffs" (2020). Theses and Dissertations--Statistics. 52.
Available for download on Sunday, September 04, 2022