• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.12
• /
• pp.4839-4842
• /
• 2014

Background: Although the Cox's proportional hazard model is the popular approach for survival analysis to investigate significant risk factors of cancer patient survival, it is not appropriate in the case of log-term disease free survival. Recently, cure rate models have been introduced to distinguish between clinical determinants of cure and variables associated with the time to event of interest. The aim of this study was to use a cure rate model to determine the clinical associated factors for cure rates of patients with breast cancer (BC). Materials and Methods: This prospective cohort study covered 305 patients with BC, admitted at Shahid Faiazbakhsh Hospital, Tehran, during 2006 to 2008 and followed until April 2012. Cases of patient death were confirmed by telephone contact. For data analysis, a non-mixed cure rate model with Poisson distribution and negative binomial distribution were employed. All analyses were carried out using a developed Macro in WinBugs. Deviance information criteria (DIC) were employed to find the best model. Results: The overall 1-year, 3-year and 5-year relative survival rates were 97%, 89% and 74%. Metastasis and stage of BC were the significant factors, but age was significant only in negative binomial model. The DIC also showed that the negative binomial model had a better fit. Conclusions: This study indicated that, metastasis and stage of BC were identified as the clinical criteria for cure rates. There are limited studies on BC survival which employed these cure rate models to identify the clinical factors associated with cure. These models are better than Cox, in the case of long-term survival.

• Journal of Technology Innovation
• /
• v.14 no.3
• /
• pp.23-42
• /
• 2006

This study investigates the determinants of output of a manufacturing firm's innovative activity (the number of patent applications) and spillover effects using a count data model. This paper attempted a negative binomial distribution In order to take into account unobserved heterogeneity. The results of our study suggested that Firm size, R&D intensity, technical network activity, and online business performance have significantly positive effects in the Korean manufacturing industry. Moreover, there are significantly positive spillover effects in the same industry sector.

• International Journal of Highway Engineering
• /
• v.11 no.2
• /
• pp.151-158
• /
• 2009

This study deals with the rear-end collisions in the rural aiea. The objectives of this study are 1) to analyze the characteristics of rear-end accidents of signalized intersections, and 2) to develop the accident models for Cheongju-Cheongwon. In pursing the above, this study gives the particular attentions to comparing the characters of urban and rural area. In this study, the dependent variables are the number of accidents and value of EPDO(equivalent property damage only), and independent variables are the traffic volumes and geometric elements. The main results analyzed are the followings. First, the statistical analyses show that the Poisson accident model using the number of accident as a dependant variable are statistically significant and the negative binomial accident model using the value of EPDO are statistically significant. Second, the independent variables of Poisson model are analyzed to be the ratio of high-occupancy vehicles, total traffic volume and the sum of exit/entry, and those of negative binomial regression are the main road width, total traffic volume and the ratio of high-occupancy vehicles. Finally, the specific independent variables to the rural area are the main road width, the ratio of high occupancy vehicle, and the sum exit/entry.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.5
• /
• pp.1087-1097
• /
• 2017

This study aims to find the best model to fit the number of insurance solicitor's turnovers of life insurance companies using count data regression models such as poisson regression, negative binomial regression, zero-inflated poisson regression, or zero-inflated negative binomial regression. Out of the four models, zero-inflated negative binomial model has been selected based on AIC and SBC criteria, which is due to over-dispersion and high proportion of zero-counts. The significant factors to affect insurance solicitor's turnover found to be a work period in current company, a total work period as financial planner, an affiliated corporation, and channel management satisfaction. We also have found that as the job satisfaction or the channel management satisfaction gets lower as channel management satisfaction, the number of insurance solicitor's turnovers increases. In addition, the total work period as financial planner has positive relationship with the number of insurance solicitor's turnovers, but the work period in current company has negative relationship with it.

• The Korean Journal of Applied Statistics
• /
• v.28 no.2
• /
• pp.231-239
• /
• 2015

A Poisson model is the first choice for counts data. Quasi Poisson or negative binomial models are usually used in cases of over (or under) dispersed data. However, these models might be unsuitable if the data consist of excessive number of zeros (zero inflated data). For zero inflated counts data, Zero Inflated Poisson (ZIP) or Zero Inflated Negative Binomial (ZINB) models are recommended to address the issue. In this paper, we further considered a situation where zero inflated data are spatially correlated. A mixed effect model with random effects that account for spatial autocorrelation is used to fit the data.

• The Korean Journal of Applied Statistics
• /
• v.22 no.2
• /
• pp.271-280
• /
• 2009

For bivariate binomial data with both intra and inter-class correlation, Danaher and Hardie (2005) proposed a bivariate beta-binomial model. However, the model is limited to the situation where the intra-class correlation is strictly positive. Thus it might be seriously inadequate for data with a negative intra-class correlation. Several authors have considered generalized binomial distributions covering a wider range of intra-class correlation which could relax the possible model restrictions imposed. Among others there are the additive/multiplicative and the beta/extended beta binomial model. In this study, bivariate models of the Sarmanov (1966) type are formed by combining each of those univariate models to take care of the inter-class correlation, and are evaluated in terms of the goodness-of-fit. As a result, B-mB and B-ebB are fitted, successfully, to real data and that B-mB, which has a wider permissible range than B-ebB for the intra-class correlation is relatively preferred.

• Journal of the Korean Society of Safety
• /
• v.30 no.1
• /
• pp.104-110
• /
• 2015

This study deals with the roundabout accidents. The goal of this study is to develop the sideswipe accident models at roundabout. In the pursuing the above, this study gives particular attentions to collecting the data of geometric structure and accidents of 54 roundabouts in Korea and developing the Poisson and negative binomial regression models. The main results are as follows. First, sideswipe accident is analyzed to be the highest frequency that is 39.5% of total accident data. Second, Poisson models which is statistically significant is developed. Finally, traffic volume per approach($X_1$), number of circulatory roadway($X_3$), operation of parking lot($X_4$) and width of circulatory roadway($X_6$) are adopted as the common variables. This study might be expected to give some implications to the accident research on the roundabout.

• International Journal of Highway Engineering
• /
• v.19 no.4
• /
• pp.53-59
• /
• 2017

PURPOSES : This study deals with traffic accidents involving trucks. The objective of this study is to develop a traffic accident model for trucks at roundabouts. METHODS : To achieve its objective, this study gives particular attention to develop appropriate models using Poisson and negative binomial regression models. Traffic accident data from 2007 to 2014 were collected from TAAS data set of road traffic authority. Thirteen explanatory variables such as geometry and traffic volume were used. RESULTS : The main results can be summarized as follows: (1) two statistically significant Poisson models (${\rho}^2=0.398$ and 0.435) were developed, and (2) the analysis revealed the common variables to be traffic volume, number of exit lanes, speed breakers, and truck apron width. CONCLUSIONS : Our modeling reveals that increasing the number of speed breakers and speed limit signs, and widening the truck apron width are important for reducing the number of truck accidents at roundabouts.

• The Journal of Information Systems
• /
• v.23 no.4
• /
• pp.169-196
• /
• 2014

This study investigates how information and communication technology (ICT) can damage intellectual property rights (IPR) in the movie industry. Utilizing a survey questionnaire to gather information about the extensive use of ICT devices, including tablet PCs and smartphones, we demonstrate how digital piracy behavior is associated with various socio-demographic characteristics. Econometrically, since a large number of people do not engage in piracy activities, we adopt a zero-inflated negative binomial model. We find that people with tablet PCs are more likely to engage in the piracy of movies from peer-to-peer (P2P) sites. In particular, when we categorize ICT devices based on whether they are portable and allow downloads, we find that people with devices equipped with both functions are most likely to engage in movie piracy.

• The Korean Journal of Applied Statistics
• /
• v.25 no.6
• /
• pp.1037-1047
• /
• 2012

The generalized linear mixed model(GLMM) is widely used in fitting categorical responses of clustered data. In the numerical approximation of likelihood function the normality is assumed for the random effects distribution; subsequently, the commercial statistical packages also routinely fit GLMM under this normality assumption. We may also encounter departures from the distributional assumption on the response variable. It would be interesting to investigate the impact on the estimates of parameters under misspecification of distributions; however, there has been limited researche on these topics. We study the sensitivity or robustness of the maximum likelihood estimators(MLEs) of GLMM for counts data when the true underlying distribution is normal, gamma, exponential, and a mixture of two normal distributions. We also consider the effects on the MLEs when we fit Poisson-normal GLMM whereas the outcomes are generated from the negative binomial distribution with overdispersion. Through a small scale Monte Carlo study we check the empirical coverage probabilities of parameters and biases of MLEs of GLMM.