• The Journal of Bigdata
• /
• v.3 no.2
• /
• pp.101-112
• /
• 2018

The purpose of the study is to predict the number of police calls using neural network which is one of the machine learning and negative binomial regression, by using the data of 112 police calls received from Chungnam Provincial Police Agency from June 2016 to May 2017. The variables which may affect the police calls have been selected for developing the prediction model : time, holiday, the day before holiday, season, temperature, precipitation, wind speed, jurisdictional area, population, the number of foreigners, single house rate and other house rate. Some variables show positive correlation, and others negative one. The comparison of the methods can be summarized as follows. Neural network has correlation coefficient of 0.7702 between predicted and actual values with RMSE 2.557. Negative binomial regression on the other hand shows correlation coefficient of 0.7158 with RMSE 2.831. Neural network has low interpretability, but an excellent predictability compared with the negative binomial regression. Based on the prediction model, the police agency can do the optimal manpower allocation for given values in the selected variables.

• Journal of the Korean Statistical Society
• /
• v.21 no.2
• /
• pp.167-177
• /
• 1992

In this paper, we investigate the limiting distribution of $M_n = max (X_1, X-2, \cdots, X_n)$ in the infinite moving average process ${X_t = \sum c_i Z_{t-i}}$ generated from i.i.d. negative binomial variables $Z_i$'s. While no limit result is possible, nonetheless asymptotic bounds are derived. We also present the tail behavior of $X_t$, i.e., weighted sum of i.i.d. random variables. This continues a study made by Rootzen (1986) for discrete innovation sequences.

• Communications for Statistical Applications and Methods
• /
• v.19 no.6
• /
• pp.761-770
• /
• 2012

We frequently encounter outcomes of count that have extra variation. This paper considers several alternative models for overdispersed count responses such as a quasi-Poisson model, zero-inflated Poisson model and a negative binomial model with a special focus on a generalized linear mixed model. We also explain various goodness-of-fit criteria by discussing their appropriateness of applicability and cautions on misuses according to the patterns of response categories. The overdispersion models for counts data have been explained through two examples with different response patterns.

• Journal of the Korean Statistical Society
• /
• v.15 no.1
• /
• pp.71-78
• /
• 1986

In this paper, multivariate discrete distribution is dealt with, where a set of r distinct counts are misreported as another set of r counts. First, the variance for the one variable marginal case is expressed in the form of an inverted parabola. Next, for the multivariate negative binomial case, elements of the covariance matrix are evaluated with reference to asymptotic distributions. Finally, for the same case of multivariate negative binomial, Bayesian estimates of the parameters and of the modification rates are provided.

• Ocean and Polar Research
• /
• v.36 no.3
• /
• pp.289-302
• /
• 2014

This paper aims to evaluate the indirect economic damages to anglers of the marine recreational charter caused by marine pollution associated with the Herbei Spirit vessel, which spilled 12,547 kl of crude oil in Taean coastal areas in December 2007. In order to evaluate the indirect cost to anglers of the charter fishing, consumer surplus for charter fishing is estimated using a Poisson model (PM), a negative binomial model (NBM), a truncated Poisson model (TPM), and a truncated negative binomial model (TNBM), which account for the characteristics of count data (non-negative discrete data), for individual travel cost method (ITCM). Because of over-dispersion problem in PM and TPM, NBM and TNBM are considered to be more appropriate statistically. All parameters such as income, fishing careers, travel cost and catch that are estimated are statistically significant and theoretically valid. Based on TNBM results, consumer surplus per trip and per person was estimated to be 277 thousand won, total consumer surplus per person and per year about 2.3 million won, and the marginal effect of consumer surplus on % changes in catch rate is about 33 thousand won. The consumer surplus was converted into total indirect economic damages for aggregation which are evaluated to be 125 billion won, reflecting the number of anglers and damage rate.

• Ocean and Polar Research
• /
• v.30 no.2
• /
• pp.141-148
• /
• 2008

This paper aims at estimating consumer surplus for recreational sea fishing in Tongyeong coastal area using individual travel cost method. A Poisson model (PM), a negative binomial model (NBM), a truncated Poisson model (TPM), and a truncated negative binomial model (TNBM) are applied for individual travel cost method in order to account characteristics of count data (non-negative discrete data.) The survey was conducted for 462 inshore anglers using personal interview method in Tongyeong during July and October 2007. Respondents were asked about how often they do fishing, travel costs, catch, income, and so on. Because of over-dispersion problem in PM and TPM, NBM and TNBM were considered to be more appropriate statistically. All parameters estimated are statistically significant and theoretically valid. As the results based on TNBM, consumer surplus per trip was estimated to be 183,486 won, total consumer surplus per person and per year 3,399,658 won, and the marginal effect of consumer surplus on % changes in catch rate is 185,372 won.

• Journal of the Korean Data and Information Science Society
• /
• v.19 no.4
• /
• pp.1327-1334
• /
• 2008

Mixed effect binomial regression models are widely used for analysis of correlated count data in which the response is the result of a series of one of two possible disjoint outcomes. In this paper, we consider kernel extensions with nonparametric fixed effects and parametric random effects. The estimation is through the penalized likelihood method based on kernel trick, and our focus is on the efficient computation and the effective hyperparameter selection. For the selection of hyperparameters, cross-validation techniques are employed. Examples illustrating usage and features of the proposed method are provided.

• Proceedings of the Korean Society for Quality Management Conference
• /
• 1998.11a
• /
• pp.198-207
• /
• 1998

The p chart approximate to the normal distribution has a difficulty to analyze the process condition precisely when the negative LCL is occurred. Furthermore, the probability of Type I error increases compared with using its original binomial distribution. For a long time the p chart has been used as approximated to the normal distribution because of its easy use. However, it becomes rapid and convenient to calculate the binomial distribution through the development of computer and software, so it is strongly suggested to use the binomial distribution determining control limits to reduce the probability of Type I error. In this study, I suggest that the control limits can be designed in use of binomial distribution and they can be utilized without special software by illustrating the certain work for establishing p-chart with the commercial one(EXCEL).

• Journal of Korean Academy of Nursing
• /
• v.42 no.6
• /
• pp.819-832
• /
• 2012

Purpose: This study was conducted to identify risk factors that influence the probability and severity of elder abuse in community-dwelling older adults. Methods: This study was a cross-sectional descriptive study. Self-report questionnaires were used to collect data from community-dwelling Koreans, 65 and older (N=416). Logistic regression, negative binomial regression and zero-inflated negative binomial regression model for abuse count data were utilized to determine risk factors for elder abuse. Results: The rate of older adults who experienced any one category of abuse was 32.5%. By zero-inflated negative binomial regression analysis, the experience of verbal-psychological abuse was associated with marital status and family support, while the experience of physical abuse was associated with self-esteem, perceived economic stress and family support. Family support was found to be a salient risk factor of probability of abuse in both verbal-psychological and physical abuse. Self-esteem was found to be a salient risk factor of probability and severity of abuse in physical abuse alone. Conclusion: The findings suggest that tailored prevention and intervention considering both types of elder abuse and target populations might be beneficial for preventative efficiency of elder abuse.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.14 no.6
• /
• pp.91-99
• /
• 2015

This study intends to build a traffic accident predictive model considering road geometrics, traffic and enviromental characteristics and identify the relationship of 4-legs intersection accidents in Seoul and Busan metropolitan area. The RPNB(Random Parameter Negative Binomial) model shows improvement over the fixed NB(Negative Binomial) and out of 53 variables, 10 variables (main road number of lane, main road vehicle traffic volume(left), minor road vehicle traffic volume(right), main road drive restriction, minor road sight distance, minor road median strip, minor road speed limit, minor road speed restriction) showed to have significant variables affecting traffic accident occurrences in 4-legs signilized intersections. Also, among 10 significant variables, 2 variables(minor road sight distance, minor road speed restriction) found to be random parameters.