A compound negative binomial distribution with mutative termination conditions based on a change point

被引:24
作者
Wang, Xiaoyue [1 ]
Zhao, Xian [1 ]
Sun, Jinglei [1 ]
机构
[1] Beijing Inst Technol, Sch Management & Econ, Beijing 100081, Peoples R China
基金
中国国家自然科学基金;
关键词
Compound random variable; Change point; Finite Markov chain imbedding approach; Phase-type distribution; PHASE-TYPE DISTRIBUTIONS; WEIGHTED DISTRIBUTIONS; ORDER K; MODEL;
D O I
10.1016/j.cam.2018.11.009
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
In this paper, an extended negative binomial distribution called NBM is proposed by considering mutative termination conditions based on a change point. If the condition A is satisfied before the occurrence of the change point, the trial is terminated according to condition A. Otherwise, if the condition A does not happen and the change point is satisfied, the termination condition of the trial changes from condition A to condition B. We consider the conditions under which the resulting distribution can be degenerated to the existing negative binomial distributions and other new negative binomial distributions. The finite Markov chain imbedding approach is employed to derive the new negative binomial distribution (NBM) and to obtain the related probabilistic indexes. Furthermore, we study the distribution of the compound negative binomial distribution with mutative termination conditions based on a change point (CNBM) by means of phase-type representations. (C) 2018 Elsevier B.V. All rights reserved.
引用
收藏
页码:237 / 249
页数:13
相关论文
共 31 条
[1]  
[Anonymous], 2014, Fundamentals of matrix-analytic methods
[2]   A weighted negative binomial Lindley distribution with applications to dispersed data [J].
Bakouch, Hassan S. .
ANAIS DA ACADEMIA BRASILEIRA DE CIENCIAS, 2018, 90 (03) :2617-2642
[3]   NEGATIVE BINOMIAL MODEL WITH VARYING ELEMENT SIZES [J].
BISSELL, AF .
BIOMETRIKA, 1972, 59 (02) :435-441
[4]   A generalized negative binomial distribution based on an extended Poisson process [J].
Bueno Salasar, Luis Ernesto ;
Leite, Jose Galvao ;
Louzada-Neto, Francisco .
BRAZILIAN JOURNAL OF PROBABILITY AND STATISTICS, 2010, 24 (01) :91-99
[5]   Discrete time shock models involving runs [J].
Eryilmaz, Serkan .
STATISTICS & PROBABILITY LETTERS, 2015, 107 :93-100
[6]   Compound Markov negative binomial distribution [J].
Eryilmaz, Serkan .
JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS, 2016, 292 :1-6
[7]   Geometric distribution of order k with a reward [J].
Eryilmaz, Serkan .
STATISTICS & PROBABILITY LETTERS, 2014, 92 :53-58
[8]   Predicting elderly patient length of stay in hospital and community care using a series of conditional Coxian phase-type distributions, further conditioned on a survival tree [J].
Gordon, Andrew S. ;
Marshall, Adele H. ;
Zenga, Mariangela .
HEALTH CARE MANAGEMENT SCIENCE, 2018, 21 (02) :269-280
[9]   A new generalization of the negative binomial distribution [J].
Gupta, RC ;
Ong, SH .
COMPUTATIONAL STATISTICS & DATA ANALYSIS, 2004, 45 (02) :287-300
[10]   Generalized binomial and negative binomial distributions of order k by the l-overlapping enumeration scheme [J].
Inoue, K ;
Aki, S .
ANNALS OF THE INSTITUTE OF STATISTICAL MATHEMATICS, 2003, 55 (01) :153-167