• Communications for Statistical Applications and Methods
• /
• v.27 no.1
• /
• pp.37-46
• /
• 2020

Recurrent event data frequently occur in clinical studies, demography, engineering reliability and so on (Cook and Lawless, The Statistical Analysis of Recurrent Events, Springer, 2007). Sometimes, two or more different but related type of recurrent events may occur simultaneously. In this study, our interest is to estimate the covariate effect on bivariate recurrent event times with zero inflations. Such zero inflation can be related with susceptibility. In the context of bivariate recurrent event data, furthermore, such susceptibilities may be different according to the type of event. We propose a joint model including both two intensity functions and two cure rate functions. Bivariate frailty effects are adopted to model the correlation between recurrent events. Parameter estimates are obtained by maximizing the likelihood derived under a piecewise constant hazard assumption. According to simulation results, the proposed method brings unbiased estimates while the model ignoring cure rate models gives underestimated covariate effects and overestimated variance estimates. We apply the proposed method to a set of bivariate recurrent infection data in a study of child patients with leukemia.

• The Korean Journal of Applied Statistics
• /
• v.25 no.4
• /
• pp.621-632
• /
• 2012

Recurrent event data can be easily found in longitudinal studies such as clinical trials, reliability fields, and the social sciences; however, there are a few observations that disappear temporarily in sight during the follow-up and then suddenly reappear without notice like the Young Traffic Offenders Program(YTOP) data collected by Farmer et al. (2000). In this article we focused on inference for a cumulative mean function of the recurrent event data with these incomplete observation gaps. Defining a corresponding risk set would be easily accomplished if we know the exact intervals where the observation gaps occur. However, when they are incomplete (if their starting times are known but their terminating times are unknown) we need to estimate a distribution function for the terminating times of the observation gaps. To accomplish this, we treated them as interval-censored and then estimated their distribution using the EM algorithm proposed by Turnbull (1976). We proposed a nonparametric estimator for the cumulative mean function and also a nonparametric test to compare the cumulative mean functions of two groups. Through simulation we investigated the finite-sample performance of the proposed estimator and proposed test. Finally, we applied the proposed methods to YTOP data.

• Journal of the Korean Data and Information Science Society
• /
• v.25 no.2
• /
• pp.327-336
• /
• 2014

Recurrent event data occurs when a subject experiences same type of event repeatedly and is found in various areas such as the social sciences, Economics, medicine and public health. To analyze recurrent event data either a total time or a gap time is adopted according to research interest. In this paper, we analyze recurrent event data with incomplete observation gap using a gap time scale. That is, some subjects leave temporarily from a study and return after a while. But it is not available when the observation gaps terminate. We adopt an interval censoring mechanism for estimating the termination time. Furthermore, to model the association among gap times of a subject, a frailty effect is incorporated into a model. Programs included in Survival package of R program are implemented to estimate the covariate effect as well as the variance of frailty effect. YTOP (Young Traffic Offenders Program) data is analyzed with both proportional hazard model and a weibull regression model.

• The Korean Journal of Applied Statistics
• /
• v.23 no.4
• /
• pp.715-724
• /
• 2010

Recurrent event data occurs when a subject experience the event of interest several times and has been found in biomedical studies, sociology and engineering. Several diverse approaches have been applied to analyze the recurrent events (Cook and Lawless, 2007). In this study, we analyzed the YTOP(Young Traffic Offenders Program) dataset which consists of 192 drivers with conviction dates by speeding violation and traffic rule violation. We consider a subject-specific effect, frailty, to reflect the individual's driving behavior and extend to time-varying frailty effect. Another feature of this study is about the redefinition of risk set. During the study, subject may be under suspension and this period is regarded as non-risk period. Thus the risk variables are reformatted according to suspension and termination time.

• Communications for Statistical Applications and Methods
• /
• v.20 no.4
• /
• pp.283-290
• /
• 2013

Subjects can experience two types of recurrent events in a longitudinal study. In addition, there may exist intermittent dropouts that results in repeated observation gaps during which no recurrent events are observed. Therefore, theses periods are regarded as non-risk status. In this paper, we consider a special case where information on the observation gap is incomplete, that is, the termination time of observation gap is not available while the starting time is known. For a statistical inference, incomplete termination time is incorporated in terms of interval-censored data and estimated with two approaches. A shared frailty effect is also employed for the association between two recurrent events. An EM algorithm is applied to recover unknown termination times as well as frailty effect. We apply the suggested method to young drivers' convictions data with several suspensions.

• Journal of the Korean Data and Information Science Society
• /
• v.25 no.5
• /
• pp.1117-1125
• /
• 2014

In a longitudinal study, subjects can experience same type of events repeatedly. Also, there may exist intermittent dropouts resulting in repeated observation gaps during which no recurrent events are observed. Furthermore, when such observation gaps have incomplete forms caused by the unknown termination times of observation gaps, ordinary approaches result in biased estimates. In this study, we investigate the effect of ignoring observation gaps and propose methods to overcome this problem. For estimating the distribution of unknown termination times, an interval-censored mechanism is applied and two cases are considered. Simulation studies are carried out to evaluate the performance of the proposed method. Conviction data of young drivers with several suspensions are analyzed to illustrate the suggested approach.

• Industry Promotion Research
• /
• v.5 no.1
• /
• pp.1-11
• /
• 2020

The purpose of this study is to analyze the industrial accidents data with survival model. EDA approach is used to explore the relationship between two variables and among three variables for the past 10 years of industrial accidents data. Survival models are also tried. Survival curve drops more rapidly for the business with fewer employees as time goes by. Industrial accidents occur more often as the total number of industrial accidents gets larger and as the number of employees gets smaller. Agriculture, fishing and forestry have a higher level of industrial accidents than construction while service industry and 'transportation·storage and telecommunication' have a fewer number of industrial accidents than construction. Korea Safety and Health Agency's and Ministry of Employment and Labor's involvement were not effective but Civilian's was. Recurrent event data analysis reveals all most the same result as for non-recurrent data analysis.

• The Korean Journal of Applied Statistics
• /
• v.25 no.5
• /
• pp.775-783
• /
• 2012

Several statistical models for bivariate poisson data are suggested and used to analyze 2011 K-league data. Our interest is composed of two purposes: The first purpose is to exploit potential attacking and defensive abilities of each team. Particular, a bivariate poisson model with diagonal inflation is incorporated for the estimation of draws. A joint model is applied to estimate an association between poisson distribution and probability of draw. The second one is to investigate causes on scoring time of goals and a regression technique of recurrent event data is applied. Some related future works are suggested.

• Communications for Statistical Applications and Methods
• /
• v.21 no.5
• /
• pp.387-393
• /
• 2014

Recurrent gap times are analyzed with diverse methods under several assumptions such as a marginal model or a frailty model. Several resampling techniques have been recently suggested to estimate the covariate effect; however, these approaches can be applied with a time-fixed covariate. According to simulation results, these methods result in biased estimates for a time-varying covariate which is often observed in a longitudinal study. In this paper, we extend a resampling method by incorporating new weights and sampling scheme. Simulation studies are performed to compare the suggested method with previous resampling methods. The proposed method is applied to estimate the effect of an educational program on traffic conviction data where a program participation occurs in the middle of the study.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.15 no.3
• /
• pp.465-472
• /
• 2020

Recently, various deep neural networks based methods have been proposed for audio event detection. In this study, we improved the performance of audio event detection by adopting an attention approach to a baseline CRNN. We applied context gating at the input of the baseline CRNN and added an attention layer at the output. We improved the performance of the attention based CRNN by using the audio data of strong labels in frame units as well as the data of weak labels in clip levels. In the audio event detection experiments using the audio data from the Task 4 of the DCASE 2018/2019 Challenge, we could obtain maximally a 66% relative increase in the F-score in the proposed attention based CRNN compared with the baseline CRNN.