• Communications for Statistical Applications and Methods
• /
• v.18 no.6
• /
• pp.733-739
• /
• 2011

The linear absolute shrinkage and selection operator(Lasso) method improves the low prediction accuracy and poor interpretation of the ordinary least squares(OLS) estimate through the use of $L_1$ regularization on the regression coefficients. However, the Lasso is not robust to outliers, because the Lasso method minimizes the sum of squared residual errors. Even though the least absolute deviation(LAD) estimator is an alternative to the OLS estimate, it is sensitive to leverage points. We propose a robust Lasso estimator that is not sensitive to outliers, heavy-tailed errors or leverage points.

• Journal of the Korean Statistical Society
• /
• v.31 no.1
• /
• pp.77-91
• /
• 2002

This paper considers a functional regression model with truncated errors in explanatory variables. We show that the ordinary least squares (OLS) estimators produce bias in regression parameter estimates under misspecified models with ignored errors in the explanatory variable measurements, and then propose methods for analyzing the functional model. Fully parametric frequentist approaches for analyzing the model are intractable and thus Bayesian methods are pursued using a Markov chain Monte Carlo (MCMC) sampling based approach. Necessary theories involved in modeling and computation are provided. Finally, a simulation study is given to illustrate and examine the proposed methods.

• Communications for Statistical Applications and Methods
• /
• v.8 no.3
• /
• pp.603-613
• /
• 2001

A test based on score statistic is derived for detecting homoscedasticity of errors in unbalanced random effects linear model. A small simulation study is performed to investigate the finite sample behaviour of the test statistic which is known to have an asymptotic chi-square distribution under the null hypothesis.

• Journal of Veterinary Clinics
• /
• v.27 no.2
• /
• pp.170-174
• /
• 2010

With the ease availability of statistical software and powerful computers the application of statistical methods in domestic veterinary journals is on the increase. In parallel with this benefit, statistical errors are not uncommon even in renowned scientific and medical journals. These errors may lead to misinterpretation of the data, thereby, subjected to faulty conclusions. A systematic review of articles published in 8 issues of the Journal of Veterinary Clinics during 2006-2007 was performed to assess the statistical methodology and reporting. Ninety-four (72.9%) articles of the 129 original articles screened included any inferential statistical analysis in the article, including comparison of 3 or more groups (53 or 56.4%), comparison of independent 2 groups (40 or 42.6%), and paired t-test (9 or 9.6%) in order. Of the 94 articles in which statistical analysis was done 62 (or 66.0%) had at least 1 statistical error. Errors included failure to apply or incorrectly applying independent Student's t-test for paired data or vice versa, inappropriate use of t-test for more than 3 groups and failure in chi-square test to consider continuity-correction for small expected frequencies. The common errors in ANOVA were failure to validate assumption of the test, inappropriate post-hoc multiple-comparison and incorrect assumption of independence of data in repeated measures design. Reporting errors included failure to state statistical methods and failure to state specific test if more than 1 test was done. It is suggested that an editorial effort would be necessary to achieve the improvement of appropriate statistical procedures through the publication of statistical guidelines to author(s).

• Journal of the Korean Data and Information Science Society
• /
• v.18 no.2
• /
• pp.419-427
• /
• 2007

Recently, Korean National Statistical Office submits the results of population projections for Korea from 1960 to 2050 year. The purpose of this paper is to suggest the reasonable assumptions for the survey of population, and then to detect the errors of the surveyed population (1960-2005) on Korean Information Statistical System.

• MALSORI
• /
• no.39
• /
• pp.53-71
• /
• 2000

With the 130 patients who were diagnosed as functional articulation disorders with no physical problems, a statistical study was performed to investigate their background information and phonological processes of errors. The results are as follows: (1) Males showed higher prevalence than females, and 5-year-old-patients showed the highest in age. (2) Most patients showed errors of 2~5 phonemes (3) The most frequent errors were found in plosives and alveolar sounds, and the most frequent phonological process of errors in the aspects of manner and place of articulation were stop-assimilations and alveolar assimilations, respectively.

• Asia pacific journal of information systems
• /
• v.7 no.2
• /
• pp.77-102
• /
• 1997

The use of statistical methods in the MIS research is increasing. However, there has been a meager attempt to critically evaluate the use statistical methods in MIS research papers. The review of 33G papers published in MIS Quarterly from Volume 1 to 14 resulted in three findings; (1) the portion of empirical research has been gradually increasing compared with non-empirical research, (2) univariate parametric statistical methods are most popular among MIS researchers, (3) researchers do not comply with the writing code of scientific research. This paper disscusses what errors MIS researchers might commit in using statistical research methods and how to prevent those errors in each of three stages of research, research design phase, statistical inference phase, and interpretation phase.

• International Journal of Computer Science & Network Security
• /
• v.23 no.9
• /
• pp.157-161
• /
• 2023

Reliability is one of the computable quality features of the software. To assess the reliability the software reliability growth models(SRGMS) are used at different test times based on statistical learning models. In all situations, Tradational time-based SRGMS may not be enough, and such models cannot recognize errors in small and medium sized applications.Numerous traditional reliability measures are used to test software errors during application development and testing. In the software testing and maintenance phase, however, new errors are taken into consideration in real time in order to decide the reliability estimate. In this article, we suggest using the Weibull model as a computational approach to eradicate the problem of software reliability modeling. In the suggested model, a new distribution model is suggested to improve the reliability estimation method. We compute the model developed and stabilize its efficiency with other popular software reliability growth models from the research publication. Our assessment results show that the proposed Model is worthier to S-shaped Yamada, Generalized Poisson, NHPP.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.5 no.4
• /
• pp.281-285
• /
• 2005

Learning and evolving are two basics for data mining. As compared with classical learning theory based on objective function with minimizing training errors, the recently evolutionary computing has had an efficient approach for constructing optimal model without the minimizing training errors. The global search of evolutionary computing in solution space can settle the local optima problems of learning models. In this research, combining co-evolving algorithm into statistical learning theory, we propose an co-evolutionary computing for statistical learning theory for overcoming local optima problems of statistical learning theory. We apply proposed model to classification and prediction problems of the learning. In the experimental results, we verify the improved performance of our model using the data sets from UCI machine learning repository and KDD Cup 2000.

• Communications for Statistical Applications and Methods
• /
• v.25 no.2
• /
• pp.131-157
• /
• 2018

We propose a robust event date model to estimate the date of a target event by a combination of individual dates obtained from archaeological artifacts assumed to be contemporaneous. These dates are affected by errors of different types: laboratory and calibration curve errors, irreducible errors related to contaminations, and taphonomic disturbances, hence the possible presence of outliers. Modeling based on a hierarchical Bayesian statistical approach provides a simple way to automatically penalize outlying data without having to remove them from the dataset. Prior information on individual irreducible errors is introduced using a uniform shrinkage density with minimal assumptions about Bayesian parameters. We show that the event date model is more robust than models implemented in BCal or OxCal, although it generally yields less precise credibility intervals. The model is extended in the case of stratigraphic sequences that involve several events with temporal order constraints (relative dating), or with duration, hiatus constraints. Calculations are based on Markov chain Monte Carlo (MCMC) numerical techniques and can be performed using ChronoModel software which is freeware, open source and cross-platform. Features of the software are presented in Vibet et al. (ChronoModel v1.5 user's manual, 2016). We finally compare our prior on event dates implemented in the ChronoModel with the prior in BCal and OxCal which involves supplementary parameters defined as boundaries to phases or sequences.