Finite-state, discrete-parameter Markov chains are used to provide a model of the population of software use to support statistical testing of software. Once a Markov chain usage model has been constructed, any number of statistically typical tests can be obtained from the model. Markov mathematics can be applied to obtain values, such as the long run probabilities, that provide information for test planning and analysis of test results. Because Markov chain usage models of industrial-sized systems are often very large, the time and memory required to compute the long run probabilities can be prohibitive. This thesis describes a procedure for automatically decomposing a large Markov chain model 'into several smaller models from which the original model's long run probabilities can be calculated. The procedure supports both parallel processing to reduce the elapsed time, and sequential processing to reduce memory requirements.
Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:honorstheses1990-2015-1209 |
Date | 01 January 2000 |
Creators | Pandya, Chirag |
Publisher | STARS |
Source Sets | University of Central Florida |
Language | English |
Detected Language | English |
Type | text |
Source | HIM 1990-2015 |
Page generated in 0.0023 seconds