Research Reports from the Department of Operations
Document Type
Dissertation
Publication Date
1-1-1986
Abstract
In many areas of research, multivariate data are collected from a subject repeatedly over time. One problem associated with these data is the classification of the subject into one of two populations. If the trends of each population have the same functional forms (polynomials), it was shown that the Regression-Discriminant (RD) procedure yields smaller error rates in general (Browdy (1978)). However, the RD procedure is not appropriate when the population trends have different functional forms from each other. The Successive Bayesian procedure is designed in this dissertation to classify time-dependent data when the population trends have different functional forms and also when there are many observations over time for each variable. The linear programming procedure, a nonparametric approach to classify time-dependent data in univariate case, is also designed in this dissertation.
Keywords
Operations research, Multivariate analysis, Time-series analysis, Discriminant analysis, Bayesian statistical decision theory, Linear programming
Publication Title
Dissertation/Technical Memorandums from the Department of Operations, School of Management, Case Western Reserve University
Issue
Technical memorandum no. 570 ; Submitted in partial fulfillment of the requirements for the Degree of Doctor of Philosophy.
Rights
This work is in the public domain and may be freely downloaded for personal or academic use
Recommended Citation
Lee, Jung Jin, "Classification of Time-Dependent Observations with Different Population Trends" (1986). Research Reports from the Department of Operations. 73.
https://commons.case.edu/wsom-ops-reports/73