Abstract
In this paper we describe a procedure for detecting a systematic change in parameter for a sequence of binomial variables. The procedure is based on a goodness of fit argument. Tests for an unknown change point are given. The procedure is found to be appropriate to problems in which the data series has been subject to a single discrete change in binomial parameter or where there have been cumulative changes in binomial parameter, before or after an unknown point.
Original language | English |
---|---|
Title of host publication | The 12th International Conference on Applied Stochastic Models and Data Analysis (ASMDA 2007) |
Editors | Christos H Skiadas, N Limnios, G Saporta |
Publisher | Birkhäuser Boston |
Publication status | Published - Jun 2007 |
Event | The 12th International Conference on Applied Stochastic Models and Data Analysis (ASMDA 2007) - Chania, Crete, Greece Duration: 29 May 2007 → 1 Jun 2007 |
Conference
Conference | The 12th International Conference on Applied Stochastic Models and Data Analysis (ASMDA 2007) |
---|---|
City | Chania, Crete, Greece |
Period | 29/05/07 → 1/06/07 |
Keywords
- Binomial parameter, Change point, Goodness of fit