• Journal of environmental and Sanitary engineering
• /
• v.23 no.4
• /
• pp.83-93
• /
• 2008

The non-linear least squares model(NLSM) has long been the standard technique used by hydrologists for constructing rating curves. The reasons for its adaptation are vague, and its appropriateness as a method of describing discharge measurement uncertainty has not been well investigated. It is shown in this paper that the classical method of NLSM can model only a very limited class of variance heterogeneity. Furthermore, this lack of flexibility often leads to unaccounted heteroscedasticity, resulting in dubious values for the rating curve parameters and estimated discharge. By introducing a heteroscedastic maximum likelihood model(HMLM), the variance heterogeneity is treated more generally. The maximum likelihood model stabilises the variance better than the NLSM approach, and thus is a more robust and appropriate way to fit a rating curve to a set of discharge measurements.

• Communications for Statistical Applications and Methods
• /
• v.16 no.6
• /
• pp.1055-1066
• /
• 2009

Many articles have considered a hybrid censoring scheme, which is a mixture of Type-I and Type-II censoring schemes. We introduce a double hybrid censoring scheme and derive some approximate maximum likelihood estimators(AMLEs) of the scale parameter for the half logistic distribution under the proposed double hybrid censored samples. The scale parameter is estimated by approximate maximum likelihood estimation method using two different Taylor series expansion types. We also obtain the maximum likelihood estimator(MLE) and the least square estimator(LSE) of the scale parameter under the proposed double hybrid censored samples. We compare the proposed estimators in the sense of the mean squared error. The simulation procedure is repeated 10,000 times for the sample size n = 20(10)40 and various censored samples. The performances of the AMLEs and MLE are very similar in all aspects but the MLE and LSE have not a closed-form expression, some numerical method must be employed.

• Journal of the Korean Statistical Society
• /
• v.26 no.3
• /
• pp.397-405
• /
• 1997

Recently maximum likelihood estimators using unconditional likelihood function are used for testing unit roots. When one wants to use this method the determinant term of initial values in the multivariate unconditional likelihood function produces a complicated function of the elements in the coefficient matrix and variance matrix. In this paper an approximation of the determinant term is calculated and based on this aproximation an approximated unconditional likelihood function is calculated. The approximated unconditional maximum likelihood estimators can be used to test for unit roots. When multivariate process has one unit root the limiting distribution obtained by this method and the limiting distribution using exact unconditional likelihood function are the same.

• The Korean Journal of Applied Statistics
• /
• v.6 no.2
• /
• pp.393-399
• /
• 1993

When we compute the maximum likelihood estimators of the parameters for the logistic regression models, which are useful in studying the relationship between the binary response variable and the explanatory variable, the standard error calculations are usually based on the second derivative of log-likelihood function. On the other hand, an estimator of the Fisher information motivated from the fact that the expectation of the cross-product of the first derivative of the log-likelihood function gives the Fisher information is expected to have similar asymptotic properties. These estimators of Fisher information are closely related with the iterative algorithm to get the maximum likelihood estimator. The average numbers of iterations to achieve the maximum likelihood estimator are compared to find out which method is more efficient, and the estimators of the variance from each method are compared as estimators of the asymptotic variance.

• Communications for Statistical Applications and Methods
• /
• v.20 no.3
• /
• pp.217-223
• /
• 2013

Two influence diagnostic methods for the common mean model are proposed. First, an investigation of the influence of observations according to minor perturbations of the common mean model is made by adapting the local influence method which is based on the likelihood displacement. It is well known that the maximum likelihood estimates are in general sensitive to influential observations. Case-deletions can be a candidate for detecting influential observations. However, the maximum likelihood estimators are iteratively computed and therefore case-deletions involve an enormous amount of computations. An approximation by Newton's method to the maximum likelihood estimator obtained after a single observation was deleted can reduce much of computational burden, which will be treated in this work. A numerical example is given for illustration and it shows that the proposed diagnostic methods can be useful tools.

• Communications for Statistical Applications and Methods
• /
• v.16 no.3
• /
• pp.479-485
• /
• 2009

When a part of data is unobserved the marginal likelihood of parameters given the observed data often involves analytically intractable high dimensional integral and hence it is hard to find the maximum likelihood estimate of the parameters. Simulated maximum likelihood(SML) method which estimates the marginal likelihood via Monte Carlo importance sampling and optimize the estimated marginal likelihood has been used in many applications. A key issue in SML is to find a good proposal density from which Monte Carlo samples are generated. The optimal proposal density is the conditional density of the unobserved data given the parameters and the observed data, and attempts have been given to find a good approximation to the optimal proposal density. Algorithms which adaptively improve the proposal density have been widely used due to its simplicity and efficiency. In this paper, we describe a fully adaptive algorithm which has been used by some practitioners but has not been well recognized in statistical literature, and evaluate its estimation performance and robustness via a simulation study. The simulation study shows a great improvement in the order of magnitudes in the mean squared error, compared to non-adaptive or partially adaptive SML methods. Also, it is shown that the fully adaptive SML is robust in a sense that it is insensitive to the starting points in the optimization routine.

• Journal of the Korean Data and Information Science Society
• /
• v.26 no.2
• /
• pp.313-322
• /
• 2015

A mixture distribution of a discrete uniform or degenerated distribution and two binomial distribution is proposed and a method of obtaining the maximum likelihood estimates of the parameters is provided. For the proposed model simulation studies were conducted to see performance of the maximum likelihood estimates and a mixture of a degenerated distribution and two binomial distributions was applied to fit a lecture evaluation data to show a good result.

• Communications for Statistical Applications and Methods
• /
• v.20 no.5
• /
• pp.417-425
• /
• 2013

We consider an extended version of a logarithmic series distribution and discuss the estimation of its parameters by the method of moments and the method of maximum likelihood. Test procedures are suggested to test the significance of the additional parameter of this distribution and all procedures are illustrated with the help of real life data sets. In addition, a simulation study is conducted to assess the performance of the estimators.

• Journal of the Korean Data and Information Science Society
• /
• v.17 no.4
• /
• pp.1319-1328
• /
• 2006

For multiply Type-II censored samples from two-parameter Rayleigh distribution, the maximum likelihood method does not admit explicit solutions. In this case, we propose some explicit estimators of the location and scale parameters in the Rayleigh distribution by the approximate maximum likelihood methods. We compare the proposed estimators in the sense of the mean squared error for various censored samples.

• Journal of the Korean Data and Information Science Society
• /
• v.11 no.1
• /
• pp.57-64
• /
• 2000

By assuming a Type-I censored sample, we propose the approximate maximum likelihood estimators(AMLE) of the scale and location parameters of the gamma distribution. We compare the proposed estimators with the maximum likelihood estimators(MLE) in the sense of the mean squared errors(MSE) through Monte Carlo method.