• Journal of the Korean Statistical Society
• /
• 제30권2호
• /
• pp.179-192
• /
• 2001

When a statistical model has a hierarchical structure such as multilayer perceptrons in neural networks or Gaussian mixture density representation, the model includes distribution with unidentifiable parameters when the structure becomes redundant. Since the exact structure is unknown, we need to carry out statistical estimation or learning of parameters in such a model. From the geometrical point of view, distributions specified by unidentifiable parameters become a singular point in the parameter space. The problem has been remarked in many statistical models, and strange behaviors of the likelihood ratio statistics, when the null hypothesis is at a singular point, have been analyzed so far. The present paper studies asymptotic behaviors of the maximum likelihood estimator and the Bayesian predictive estimator, by using a simple cone model, and show that they are completely different from regular statistical models where the Cramer-Rao paradigm holds. At singularities, the Fisher information metric degenerates, implying that the cramer-Rao paradigm does no more hold, and that he classical model selection theory such as AIC and MDL cannot be applied. This paper is a first step to establish a new theory for analyzing the accuracy of estimation or learning at around singularities.

• Journal of Chest Surgery
• /
• 제27권9호
• /
• pp.732-737
• /
• 1994

From January 1983 to December 1992, There were 1441 papers in the Korean Journal of Thoracic and Cardiovascular Surgery. Among these papers, 783[54.3%] were original article or clinical analysis and 652[45.2%] were case reports. A total of 319 papers contained some statistical analysis. In 150 cases[47.0%] of these 319 papers, the statistical description was insufficient. Of the correctly described papers, 115[68%] had more than one statistical error. Of course, in many cases the errors were not considered to be severe, but they were often sufficient to raise doubts about some inferences. We suggest that authors should be more careful when they describe and apply statistical methods. If possible, authors should interpret results with statistical specialists. And we also suggest that our society have more extensive statistical refereeing system. This would at least prevent the worst errors from appearing in print. The last suggestion is elementary instruction in statistical methods during preclinical training.

• Communications for Statistical Applications and Methods
• /
• 제16권3호
• /
• pp.437-447
• /
• 2009

최근에 산업현장에서의 통계전문가들에게는 여러 가지 통계분석기법을 사용한 자료 분석 외에 다양한 형태의 자료 저장장치에서 추출 또는 생성의 과정을 거쳐 분석 목적에 적합한 자료를 구성해야하는 문제에 많이 부닥치고 있다. 본 논문에서는 현재 일반적으로 사용되고 있는 여러 통계 패키지들에서 제공하고 있는 데이터 접근방식을 살펴보고 각 기능들을 비교 분석하고자 한다. 이들 방식에 대한 정확한 이해는 특히 데이터마이닝 등 대용량의 자료를 분석하고자 할 때 데이터 처리과정에서의 어려움으로 발생하는 비용과 시간을 감소시켜주어 통계전문가들이 통계분석에 더욱 많은 작업을 할애할 수 있도록 해줄 것이다.

• 응용통계연구
• /
• 제26권3호
• /
• pp.421-429
• /
• 2013

현대통계학은 많은 분야에서 사용되고 있으나 사용자들이 통계학적 개념을 이해하는데 어려움을 겪고 있다. 한편으로는 초등학생 때부터 줄기잎 그림이나, 비율자료의 원그림 등은 물론이고 평균과 같은 기술통계를 배우고 있다. 초등학교 고학년이나 중학교 저학년 학생들을 위한 직관적인 통계 패키지가 있다면 미래의 통계 사용자들인 이들 학생들이 통계적 개념을 이해하는데 많은 도움이 될 것이라고 생각하여 직관에 기초한 통계 패키지를 개발하였다.

• 한국수학교육학회지시리즈C:초등수학교육
• /
• 제12권2호
• /
• pp.133-143
• /
• 2009

통계는 삶의 문제에서 출발하여 주어진 문제를 해결하기 위해 통계문제로 번안되고, 자료가 정의되고 수집되어 적절하게 통계를 활용하여 원문제의 해결을 해야 한다. 그러므로 통계교육은 이러한 연장선상에서 유기적이고 체계적으로 이루어져야 한다. 이 논문에서는 초등 통계 영역이 가지는 문제점들로 교육과정상의 문제점, 교과서 구성 측면에서의 문제점, 수업 운영의 문제점, 실생활과의 연계 부족, 다른 영역과의 연계성의 문제, 통계에 대한 인식의 문제 등을 살펴보았다. 이러한 문제점의 해결방안에 대해 탐구하여 초등통계교육의 개선을 위한 시사점을 제공하는데 그 목적이 있다.

• 한국수학교육학회지시리즈A:수학교육
• /
• 제50권2호
• /
• pp.185-212
• /
• 2011

The purpose of this study is to analyze whether spreadsheet-used instruction can improve statistical thinking ability and attitude and also to identify what characteristics of statistical thinking is constructed. For this study, a subject of 2 classes were randomly selected among the 12 classes of the 11th grader in D high school and designated one class as the experimental group and the other class as the control group. Eight hours of the spread sheet-used instruction and the traditional textbook-oriented instruction had been carried out in each class. The research findings are as follows. First, the spread sheet-used instruction is shown to be more effective in enhancing statistical thinking than the traditional textbook-oriented instruction. Second, the spread sheet-used instruction is shown to be more effective in improving statistical attitude than the traditional textbook-oriented instruction. Third, students have shown the various characteristics of statistical thinking in the data descriptive process, data arrange-summary process, data representing process, and data analying process through the spread sheet-used instructions. Hence, the spread sheet-used instruction is recommended in teaching statistics.

• Journal of Information Science Theory and Practice
• /
• 제11권2호
• /
• pp.67-81
• /
• 2023

This paper presents a dynamic development of an ontology-based statistical information system supporting the collection, storage, processing, analysis, and the presentation of statistical knowledge at the national scale. To accomplish this, we propose a data mining technique to dynamically collect data relating to citizens from publicly available data sources; the collected data will then be structured, classified, categorized, and integrated into an ontology. Moreover, an intelligent platform is proposed in order to generate quantitative and qualitative statistical information based on the knowledge stored in the ontology. The main aims of our proposed system are to digitize administrative tasks and to provide reliable statistical information to governmental, economic, and social actors. The authorities will use the ontology-based statistical information system for strategic decision-making as it easily collects, produces, analyzes, and provides both quantitative and qualitative knowledge that will help to improve the administration and management of national political, social, and economic life.

• 스마트미디어저널
• /
• 제3권3호
• /
• pp.26-35
• /
• 2014

통계그래픽스는 정량적인 데이터를 이용하여 정보 분석, 추출, 시각화의 과정을 거쳐 정확한 정보전달과 효과적인 이해를 위해 사용자 인지측면에 초점을 둔 디자인 분야이다. 이러한 통제그래픽스에 빅데이터의 구성요소들 내포하게 될 경우 빅데이터 통제그래픽스라고 할 수 있다. 통계그래픽스에서 시각적 요소는 인지부분에 대한 오류를 줄이고 성공적으로 정보를 전달하기 위해 사용되어야 하지만, 빅데이터 통계그래픽스에서는 방대한 데이터로 인해 시각적 요소가 오히려 인지적 방해를 일으키고 있다. 본 연구는 빅데이터 통계 그래픽스에서 나타날 수 있는 인지적 방해요소를 도출하여 제시하는 것을 목적으로 한다. 빅데이터의 통계그래픽스의 유형을 구조적 형태를 바탕으로 '네트워크 유형', '세그먼트 유형', '혼합유형' 세 가지로 분류하였고, 그에 따른 특징들을 탐색하였다. 특히, 빅데이터 통계그래픽스에서 시각적 주요요소를 기반으로 시각화의 고도화 시 나타날 수 있는 인지적 방해요소를 '다차원 범례', '다양한 색채', '정보의 중첩', '서체의 가독성' 네 가지로 도출하여 제시하였다.

• 천문학회지
• /
• 제40권4호
• /
• pp.99-106
• /
• 2007

Statistical analyses were performed to investigate the relative success and accuracy of daily maximum X-ray flux (MXF) predictions, using both multilinear regression and autoregressive time-series prediction methods. As input data for this work, we used 14 solar activity parameters recorded over the prior 2 year period (1989-1990) during the solar maximum of cycle 22. We applied the multilinear regression method to the following three groups: all 14 variables (G1), the 2 so-called 'cause' variables (sunspot complexity and sunspot group area) showing the highest correlations with MXF (G2), and the 2 'effect' variables (previous day MXF and the number of flares stronger than C4 class) showing the highest correlations with MXF (G3). For the advanced three days forecast, we applied the autoregressive timeseries method to the MXF data (GT). We compared the statistical results of these groups for 1991 data, using several statistical measures obtained from a $2{\times}2$ contingency table for forecasted versus observed events. As a result, we found that the statistical results of G1 and G3 are nearly the same each other and the 'effect' variables (G3) are more reliable predictors than the 'cause' variables. It is also found that while the statistical results of GT are a little worse than those of G1 for relatively weak flares, they are comparable to each other for strong flares. In general, all statistical measures show good predictions from all groups, provided that the flares are weaker than about M5 class; stronger flares rapidly become difficult to predict well, which is probably due to statistical inaccuracies arising from their rarity. Our statistical results of all flares except for the X-class flares were confirmed by Yates' $X^2$ statistical significance tests, at the 99% confidence level. Based on our model testing, we recommend a practical strategy for solar X-ray flare predictions.

• Journal of Acupuncture Research
• /
• 제20권1호
• /
• pp.144-158
• /
• 2003

This study was carried out to investigate what kinds of statistical techniques have been used to analyze data from oriental medicine research, For study, 551 original articles which used statistical techniques in their data analysis were selected form the articles published in The journal of Korean Acupuncture & Moxibustion Society(JKAMS) between 1984 to 2002. among them, 122 articles used descriptive statistics while 429 articles used inferential statistics for data analysis. For that 429 articles, t-test (189 articles), analysis fo variance (111 articles), chi-square test (14 articles), correlation (10 articles), regression analysis (4 articles), factor analysis(5 articles), or nonparametric test (23 articles) were chose to analyze the data. Nonparametric approach has substantial power in case data do not meet the assumption of normality. This method is not only easy to use ut also provides measures of the statistical variation of nominal and ordinal scale. This study shows that more and more recent papers use nonparametric test compared to the old articles. nine different statistical software or packages (SAS, SPSS, Statview, Minitab, Sigma plot, ISP, Graphpad prism, Excel, Access) have been used in the articles published JKMAS. High level statistical techniques such as SAS, SPSS, and Statview are user friendly and used most for acupuncture and Moxibustion research. Including tables and plots in an article facilitates understanding family process data from a descriptive standpoint, minimized erroneous statistical conclusions, and clarifies theoretically important relationships among variables. Table and plots have been used 500 and 233 articles, respectively. A computer procedure is proposed and illustrated with statistical packages using SAS, SPSS, Statview and ISP.