• Journal of the Korean Data and Information Science Society
• /
• v.18 no.2
• /
• pp.301-314
• /
• 2007

The aim of this paper is to deal with a method to construct a separate fuzzy regression model with crisp input and fuzzy output data using a best response function for the center and the width of the predicted output. Also we introduce the crisp mean and variance of the predicted fuzzy value and also give some examples to compare a performance of the proposed fuzzy model with various other fuzzy regression model.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.44 no.1
• /
• pp.45-50
• /
• 2021

Machines and facilities are physically or chemically degenerated by continuous usage. The representative type of the degeneration is the wearing of tools, which results in the process mean shift. According to the increasing wear level, non-conforming products cost and quality loss cost are increasing simultaneously. Therefore, a preventive maintenance is necessary at some point. The problem of determining the maintenance period (or wear limit) which minimizes the total cost is called the 'process mean shift problem'. The total cost includes three items: maintenance cost (or adjustment cost), non-conforming cost due to the non-conforming products, and quality loss cost due to the difference between the process target value and the product characteristic value among the conforming products. In this study, we set the production volume as a decreasing function rather than a constant. Also we treat the process variance as a function to the increasing wear rather than a constant. To the quality loss function, we adopted the Cpm+, which is the left and right asymmetric process capability index based on the process target value. These can more reflect the production site. In this study, we presented a more extensive maintenance model compared to previous studies, by integrating the items mentioned above. The objective equation of this model is the total cost per unit wear. The determining variables are the wear limit and the initial process setting position that minimize the objective equation.

• The Korean Journal of Applied Statistics
• /
• v.24 no.3
• /
• pp.485-494
• /
• 2011

When the observations can take only the non-negative integer values, it is called the count data such as the numbers of car accidents, earthquakes, or insurance coverage. In general, the Poisson regression model has been used to model these count data; however, this model has a weakness in that it is restricted by the equality of the mean and the variance. On the other hand, the count data often tend to be too dispersed to allow the use of the Poisson model in practice because the variance of data is significantly larger than its mean due to heterogeneity within groups. When overdispersion is not taken into account, it is expected that the resulting parameter estimates or standard errors will be inefficient. Since coverage is the main issue for insurance, some accidents may not be covered by insurance, and the number covered by insurance may be zero. This paper considers the zero-inflated model for the count data including many zeros. The performance of this model has been investigated by using of real data with overdispersion and many zeros. The results indicate that the Zero-Inflated Negative Binomial Regression Model performs the best for model evaluation.

• Journal of Korean Society for Quality Management
• /
• v.27 no.1
• /
• pp.59-79
• /
• 1999

The fitting of a response surface model and the subsequent exploration of the response surface are usually based on the assumption that the experimental runs are carried out under homogeneous conditions. This, however, may be quite often difficult to achieve in many experiments. To control such an extraneous source of variation, the response surface design should be arranged in several blocks within which homogeneity of conditions can be maintained. In this case, when fitting a response surface model, the least squares estimates of the model's parameters and the prediction variance will generally depend on how the response surface design is blocked. That is, the choice of a blocking arrangement for a response surface design can have a considerable effect on estimating the mean response and on the size of the prediction variance. In this paper, we propose a measure for evaluating the effect of blocking of response surface designs using cuboidal regions.

• Animal Bioscience
• /
• v.34 no.4
• /
• pp.471-481
• /
• 2021

Objective: A comprehensive study was conducted to study the effects of partition of variance on accuracy of genetic parameters and genetic trends of economic traits in Vanaraja male line/project directorate-1 (PD-1) chicken. Methods: Variance component analysis utilizing restricted maximum likelihood animal model was carried out with five generations data to delineate the population status, direct additive, maternal genetic, permanent environmental effects, besides genetic trends and performance of economic traits in PD-1 chickens. Genetic trend was estimated by regression of the estimated average breeding values (BV) on generations. Results: The body weight (BW) and shank length (SL) varied significantly (p≤0.01) among the generations, hatches and sexes. The least squares mean of SL at six weeks, the primary trait was 77.44±0.05 mm. All the production traits, viz., BWs, age at sexual maturity, egg production (EP) and egg weight were significantly influenced by generation. Model four with additive, maternal permanent environmental and residual effects was the best model for juvenile growth traits, except for zero-day BW. The heritability estimates for BW and SL at six weeks (SL6) were 0.20±0.03 and 0.17±0.03, respectively. The BV of SL6 in the population increased linearly from 0.03 to 3.62 mm due to selection. Genetic trend was significant (p≤0.05) for SL6, BW6, and production traits. The average genetic gain of EP40 for each generation was significant (p≤0.05) with an average increase of 0.38 eggs per generation. The average inbreeding coefficient was 0.02 in PD-1 line. Conclusion: The population was in ideal condition with negligible inbreeding and the selection was quite effective with significant genetic gains in each generation for primary trait of selection. The animal model minimized the over-estimation of genetic parameters and improved the accuracy of the BV, thus enabling the breeder to select the suitable breeding strategy for genetic improvement.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.1
• /
• pp.155-162
• /
• 2010

In this paper we propose a doubly penalized kernel method which estimates both the mean function and the variance function simultaneously by kernel machines for heteroscedastic autoregressive data. We also present the model selection method which employs the cross validation techniques for choosing the hyper-parameters which aect the performance of proposed method. Simulated examples are provided to indicate the usefulness of proposed method for the estimation of mean and variance functions.

• Structural Engineering and Mechanics
• /
• v.56 no.5
• /
• pp.767-785
• /
• 2015

The stochastic finite element method is employed to obtain a stochastic dynamic model of angled beams subjected to impact loads when uncertain material properties are described by random fields. Using the perturbation technique in conjunction with a precise time integration method, a random analysis approach is developed for efficient analysis of random elastic waves. Formulas for the mean, variance and covariance of displacement, strain and stress are introduced. Statistics of displacement and stress waves is analyzed and effects of bend angle and material stochasticity on wave propagation are studied. It is found that the elastic wave correlation in the angled section is the most significant. The mean, variance and covariance of the stress wave amplitude decrease with an increase in bend angle. The standard deviation of the beam material density plays an important role in longitudinal displacement wave covariance.

• The Korean Journal of Applied Statistics
• /
• v.13 no.2
• /
• pp.407-414
• /
• 2000

Carroll and Ruppert(1988) analyzed the esterase assay data with regression model based on quasi-likelihood. Jung and Lee(1997) introduced a goodness-of-fit test for testing the adequacy of the quasi-likelihood and claimed that there is no gross inadequacy with the model because their test was not rejected. However, Lee and Xelder(199S)'s residual plots revealed that the model did not sufficiently reflect the increase of the variance with that of the mean. In this paper, we re-analyze the esterase assay data with the joint modelling of mean and dispersion in Lee and l\elder(1998) and evaluate the validity of the fitted model by applying the residual plots. And it is illustrated that Lee and Nelder(199S)'s restricted likelihood is more efficient in goodness-of-fit test for the dispersion model.

• Journal of Korean Society for Quality Management
• /
• v.24 no.2
• /
• pp.44-53
• /
• 1996

This paper deals with the problem of selecting optimal wear limit and initial process mean in a wear-out process. Consider a material removal processing machinery where 1)there are deffective items by linear shift in the mean of the diameter of to be processed with varing process variance and 2)there can be any failure in the machine tools or to be processed. In the previous studies, the one is analyzed by 'Model of Producing Goods' in quality control area and the other, any failure, is analyzed by 'Model of Producing Services' in reliability area. We propose a new integrated maintenance model, considering the percent defective and the failure rate. A numerical example for the model is given.

• Journal of the Korean Statistical Society
• /
• v.30 no.1
• /
• pp.77-87
• /
• 2001

Some asymptotic results on the local likelihood density estimator of Copas(1995) are derived when the locally parametric model has several parameters. It turns out that it has the same asymptotic mean squared error as that of Hjort and Jones(1996).