• Journal of the Korean Data and Information Science Society
• /
• 제14권2호
• /
• pp.177-186
• /
• 2003

The Zero-Inflated Poisson regression is a model for count data with exess zeros. When the correlated response variables are intrested, we have to extend the univariate zero-inflated regression model to multivariate model. In this paper, we study and simulate the multivariate zero-inflated regression model. A real example was applied to this model. Regression parameters are estimated by using MLE's. We also compare the fitness of multivariate zero-inflated Poisson regression model with the decision tree model.

• 품질경영학회지
• /
• 제46권3호
• /
• pp.509-522
• /
• 2018

Purpose: The purpose of this study was to propose more accurate mathematical model which can represent result of government quality assurance activity, especially corrective action and flaw. Methods: The collected data during government quality assurance activity was represented through histogram. To find out which distributions (Poisson distribution, Zero-Inflated Poisson distribution) could represent the histogram better, this study applied Pearson's correlation coefficient. Results: The result of this study is as follows; Histogram of corrective action during past 3 years and Zero-Inflated Poisson distribution had strong relationship that their correlation coefficients was over 0.94. Flaw data could not re-parameterize to Zero-Inflated Poisson distribution because its frequency of flaw occurrence was too small. However, histogram of flaw data during past 3 years and Poisson distribution showed strong relationship that their correlation coefficients was 0.99. Conclusion: Zero-Inflated Poisson distribution represented better than Poisson distribution to demonstrate corrective action histogram. However, in the case of flaw data histogram, Poisson distribution was more accurate than Zero-Inflated Poisson distribution.

• 응용통계연구
• /
• 제19권3호
• /
• pp.505-519
• /
• 2006

셀 수 있는 이산 자료(discrete count data)에 대한 분석은 여러 분야에서 활용되고 있지만 영(zero)을 과도하게 포함하고 있는 영과잉 자료는 자료의 성격상 포아송 분포를 따르지 못할 때가 있어 분석에 어려움이 따른다. Zero-Inflated Poisson(ZIP)모형은 이런 어려움을 극복하기 위하여 영에 대한 점확률을 가지는 분포와 포아송 분포를 합성하여 과도한 영과 영이 아닌 자료를 설명하는 모형이다. 설명 변수가 존재할 때는 포아송 분포 부분에서 반응변수의 평균과 공변량사이에 로그선형 연결함수를 사용한 Zero-Inflated Poisson Regression(ZIPR)모형이 사용될 수 있다. 본 논문에서는 Markov Chain Monte Carlo 기법을 이용한 ZIPR모형의 베이지안 추론방법을 제안하고, 이를 실제 구강위생 자료에 적용하며 다른 모형들과 비교한다. 그 결과 베이지안 추론 방법을 적용한 영과잉 모형의 추정오차가 다른 모형들의 추정오차보다 작았고, 예측치가 더 정확했다는 점에서 우수함을 알 수 있었다.

• 한국수학교육학회지시리즈B:순수및응용수학
• /
• 제20권1호
• /
• pp.59-70
• /
• 2013

The purpose of this study was to develop a zero-inflated Rasch (ZI-Rasch) model, a combination of the Rasch model and the ZIP model. The ZI-Rasch model was considered in this study as an appropriate alternative to the Rasch model for zero-inflated data. To investigate the relative appropriateness of the ZI-Rasch model, several analyses were conducted using PROC NLMIXED procedures in SAS under various simulation conditions. Sets of criteria for model evaluations (-2LL, AIC, AICC, and BIC) and parameter estimations (RMSE, and $r$) from the ZI-Rasch model were compared with those from the Rasch model. In the data-model fit indices, regardless of the simulation conditions, the ZI-Rasch model produced better fit statistics than did the Rasch model, even when the response data were generated from the Rasch model. In terms of item parameter ${\lambda}$ estimations, the ZI-Rasch model produced estimates similar to those of the Rasch model.

• Journal of the Korean Data and Information Science Society
• /
• 제14권1호
• /
• pp.45-53
• /
• 2003

The Zero-Inflated Poisson regression is a model for count data with exess zeros. When the reponse variables have excess zeros, it is not easy to apply the Poisson regression model. In this paper, we study and simulate the zero-inflated Poisson regression model. An real example was applied to this model. Regression parameters are estimated by using MLE's. We also compare the fitness of zero-inflated Poisson model with the Poisson regression and decision tree model.

• Journal of the Korean Data and Information Science Society
• /
• 제15권2호
• /
• pp.387-394
• /
• 2004

Zero-Inflated Poisson distribution is Poisson distribution with excess zeros. Recently defects of product hardley happen in the manufacturing process. In this case it is desirable to apply to the Zero-Inflated Poisson distribution rather than Poisson. Our target of this paper is to study the tests for changes of rate of defects after the unknown change-point. We are going to compare the powers of the two proposed tests with likelihood tests by the simulations.

• 응용통계연구
• /
• 제35권2호
• /
• pp.311-325
• /
• 2022

0의 값을 과도하게 포함하는 가산자료는 다양한 연구 분야에서 흔히 나타난다. 영과잉 모형은 영과잉 가산자료를 분석하기 위해 가장 일반적으로 사용되는 모형이다. 영과잉 모형에 대한 전통적인 베이지안 추론은 조건부 사후분포의 형태가 폐쇄형 분포로 나타나지 않아 모형 적합 과정이 용이하지 않다는 한계점이 존재했다. 그러나 최근 Pillow와 Scott (2012)과 Polson 등 (2013)이 제안한 폴랴-감마 자료확대전략으로 인해, 로지스틱 회귀모형과 음이항 회귀모형에서 깁스 샘플링을 통한 추론이 가능해지면서, 영과잉 모형에 대한 베이지안 추론이 용이해졌다. 본 논문에서는 베이지안 추론에 기반한 영과잉 음이항 회귀모형을 Min과 Agresti(2005)에서 분석된 약학 연구 자료에 적용해본다. 분석에 사용된 자료는 경시적 영과잉 가산자료로 복잡한 자료 구조를 가지고 있다. 모형 적합 과정에서는 깁스 샘플링을 통한 추론을 수행하기 위해 폴랴-감마 자료확대전략을 사용한다.

• 한국도로학회논문집
• /
• 제18권4호
• /
• pp.83-92
• /
• 2016

PURPOSES : The purpose of this study was to develop safety performance functions (SPFs) that use zero-inflated negative binomial regression models for urban intersections in central business districts (CBDs), and to compare the statistical significance of developed models against that of regular negative binomial regression models. METHODS : To develop and analyze the SPFs of intersections in CBDs, data acquisition was conducted for dependent and independent variables in areas of study. We analyzed the SPFs using zero-inflated negative binomial regression model as well as regular negative binomial regression model. We then compared the results by analyzing the statistical significance of the models. RESULTS : SPFs were estimated for all accidents and injury accidents at intersections in CBDs in terms of variables such as AADT, Number of Lanes at Major Roads, Median Barriers, Right Turn with an Exclusive Turn Lane, Turning Guideline, and Front Signal. We also estimated the log-likelihood at convergence and the likelihood ratio of SPFs for comparing the zero-inflated model with the regular model. In he SPFs, estimated log-likelihood at convergence and the likelihood ratio of the zero-inflated model were at -836.736, 0.193 and -836.415, 0.195. Also estimated the log-likelihood at convergence and likelihood ratio of the regular model were at -843.547, 0.187 and -842.631, 0.189, respectively. These figures demonstrate that zero-inflated negative binomial regression models can better explain traffic accidents at intersections in CBDs. CONCLUSIONS : SPFs that use a zero-inflated negative binomial regression model demonstrate better statistical significance compared with those that use a regular negative binomial regression model.

• 응용통계연구
• /
• 제28권2호
• /
• pp.231-239
• /
• 2015

가산자료(counts data)를 적합 하는 경우 보통 포아송 모형이 가장 먼저 고려된다. 과산포 문제가 있을 경우도 유사 포아송(quasi Poisson) 모형이나 음이항(Negative binomial) 모형으로 대부분 설명이 가능하다. 하지만, 가산자료 중에는 포아송분포를 가정한 기대 빈도 이상으로 많은 0이 관측되는 자료가 있고 이를 영과잉(Zero inflated) 가산 자료라고 부른다. 영과잉 가산자료를 설명하기 위해 영과잉 포아송(ZIP) 모형이나 영과잉 음이항(ZINB) 모형을 이용할 수 있다. 더 나아가 영과잉 가산자료가 공간상관관계까지 있을 경우 영과잉 문제뿐만 아니라 유의할 수 있는 공간효과까지 고려해야하고 이를 위해 혼합효과모형(mixed effects model)이 고려 될 수 있다. 본 연구에서 사용된 2004년 기준 부산시 남성동별 갑상선암 발생자수 자료를 이용하여, 일반선형 포아송모형, 영과잉 포아송모형, 공간 영과잉 포아송모형을 적합하여 비교해보았다.

• 응용통계연구
• /
• 제25권5호
• /
• pp.859-864
• /
• 2012

Poisson regression and negative binomial regression are usually used to analyze counting data; however, these models are unsuitable for fit zero-inflated data that contain unexpected zero-valued observations. In this paper, we review the zero-inflated regression in which Bernoulli process and the counting process are hierarchically mixed. It is known that zero-inflated regression can efficiently model the over-dispersion problem. Vuong statistic is employed to compare performances of the zero-inflated models with other standard models.