Joint distributions associated with patterns,successes and failures in a sequence of multi-state trials |
| |
Authors: | Inoue Kiyoshi |
| |
Affiliation: | (1) The Institute of Statistical Mathematics, 4-6-7 Minami-Azabu, Minato-ku, 106-8569 Tokyo, Japan |
| |
Abstract: | Let {Z t ,t≥1} be a sequence of trials taking values in a given setA={0, 1, 2,...,m}, where we regard the value 0 as failure and the remainingm values as successes. Let ε be a (single or compound) pattern. In this paper, we provide a unified approach for the study of two joint distributions, i.e., the joint distribution of the numberX n of occurrences of ε, the numbers of successes and failures inn trials and the joint distribution of the waiting timeT r until ther-th occurrence of ε, the numbers of successes and failures appeared at that time. We also investigate some distributions as by-products of the two joint distributions. Our methodology is based on two types of the random variablesX n (a Markov chain imbeddable variable of binomial type and a Markov chain imbeddable variable of returnable type). The present work develops several variations of the Markov chain imbedding method and enables us to deal with the variety of applications in different fields. Finally, we discuss several practical examples of our results. This research was partially supported by the ISM Cooperative Research Program (2002-ISM·CRP-2007). |
| |
Keywords: | Run pattern waiting time enumeration schemes Markov chain double generating function probability generating function Markov chain imbedding method transition probability matrices |
本文献已被 SpringerLink 等数据库收录! |
|