• The Journal of the Korea Contents Association
• /
• v.18 no.2
• /
• pp.468-475
• /
• 2018

The current system of making course evaluations mandatory in universities in Korea in order to supplement low participation levels is a major hindering factor for the validity of the evaluation results due to students' insincere responses. This study points out the problems in course evaluations that are in the form of surveys and instead proposes a sample course evaluation system that utilizes a sample mean to deduce a population mean, attempting to prove through simulation that this is statistically significant. Thus, small, medium, and large scale courses as well as cyber courses that were available in S University in the fall 2016 semester were set as the population, and each course's evaluation average grade was set as the population mean. Furthermore, we used the average grade and standard deviation of the same courses from the previous year in order to decide the sample number, the reliability level was set as 95%, and the margin of error was set as ${\pm}0.25$. As a result of carrying out a simulation using the data analysis tool R, all the courses (small scale, medium scale, large scale, and cyber) showed to have a reliability level close to 95% including the population mean, and consequently practicality of the representative sample course evaluation system was proven.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.4-4
• /
• 2017

수문자료의 빈도해석은 자료의 독립성(independence)와 정상성(stationarity)를 가정하여 이뤄진다. 그러나 관측 수문자료에서 비정상성 현상이 발생하고 있다는 사실이 관측되면서 수문자료에 대한 비정상성 빈도해석에 대한 필요성도 커지고 있다. 본 연구의 목적은 수문자료의 빈도해석에서 가장 널리 사용되고 있는 Gumbel 및 GEV 분포에 대한 비정상성 빈도해석 모형을 개발하는 것으로, 이를 위해 비정상성 Gumbel과 GEV 모형의 매개변수를 시간에 따라 변하는 형태로 정의하였다. 비정상성 Gumbel 및 GEV 모형의 정확도를 알아보기 위해 비정상성 모형과정상성 모형을 이용하여 Monte Carlo 모의실험을 수행하였다. 모의실험은 다양한 조건의 재현기간, 표본크기, 매개변수 조건을 고려하여 수행되었다. 그 결과 비정상성 모형의 오차는 비교적 표본크기가 클 때 가장 작은 것으로 나타났다. 또한 복잡한 매개변수의 조합을 가지는 비정상성 모형은 모두 동일한 경향성을 가질 때 가장 작은 오차를 보이는 것으로 나타났다. 비정상성 GEV 모형의 경우는 확률수문량 산정에 음(-)의 형상 매개변수가 큰 영향을 끼치는 것으로 나타났다. 또한 본 연구에서는 비정상성 조건에서 다양하게 존재하는 비정상성 모형 중 어떠한 모형이 주어진 자료에 대해 가장 적절한 모형인지 결정하기 위해 모의실험을 수행하였다. 널리 적용되고 있는 AIC, BIC, likelihood ratio test에 대해 정상성 및 비정상성 Gumbel 모형을 이용하여 모의실험을 수행한 결과, AIC가 비정상성 모형 중 적정 모형 선택에 가장 효과적인 것으로 나타났다. 개발된 비정상성 Gumbel 및 GEV 모형의 적용성을 알아보기 위해 우리나라 연최대강우 자료에 적용한 결과, 위치 매개변수에 시간항을 고려하는 Gumbel 모형이 최적모형으로 가장 많이 선택되는 것으로 나타났다. 따라서 현재 우리나라의 연최대강우자료 중 경향성이 나타나는 자료에 대해서는 위치 매개변수가 시간에 따라 변하는 특성이 가장 많이 나타나고 있는 것으로 판단된다.

• Journal of Korean Society for Quality Management
• /
• v.13 no.1
• /
• pp.77-83
• /
• 1985

This paper deals with a normality test to determine whether the data are sampled from normal population or not. In this paper the property that the mean and variance are independently distributed only for the normal distribution is used as a basis for developing a new test using the simple regression analysis. Considering the redan and variance of a random sample as independent and dependent variables, if it has not the regression relationship we conclude that the data were sampled from the normal distribution. The Monte-Carlo power study shows that the new test using the simple regression analysis has good power property relative to 6 well-known test methods for 11 distributions.

• The Korean Journal of Applied Statistics
• /
• v.8 no.1
• /
• pp.119-131
• /
• 1995

Based on Monte-Carlo simulation results we propose a robust analysis of variance procedure by utilizing trimmed mean and Winsorized variance. We deal with mainly the one-way classification case. We evaluate the empirical distribution of a pseudo-F statistic based on symmetrically Winsorized sum of squares when the population is normally distributed.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2021.01a
• /
• pp.355-356
• /
• 2021

본 논문에서는 Python 3의 Keras 모듈을 이용하여 특정 자동차에 대한 최적의 판매자권장소비자가격(MSRP)을 예측하는 시스템을 제안한다. 이 시스템은 2004년에 미국에서 시판된 428종류의 자동차에 대한 정보를 제조사, 차종, 생산지, 엔진 크기, 실린더 수, 시내 주행 시 연비, 고속도로 주행 시 연비, 마력, 차체 무게, 차체 길이의 독립변수를 사용하여 자체적으로 딥러닝한 회귀모델을 통해 특정 지표가 주어진 차량에 대해 종속변수인 판매자권장소비자가격을 예측한다. Optimizer를 adam으로, 학습률을 0.005으로 설정한 경우의 검증 MAE 값이 3842.98로 가장 낮게 산출되었고, 해당 모델의 결과는 예측값과 실제값의 오차율이 ±15% 정도 내외로 예측된 표본의 비율이 약 80.14%로 측정되었다. 위 모델은 향후 신차 가격 결정 및 중고차 시장에서 구매, 판매 결정을 돕는 등 특정 시장 내에서 다양한 자동차의 가치를 판단하기에 유용할 것으로 전망된다.

• Communications for Statistical Applications and Methods
• /
• v.15 no.3
• /
• pp.353-365
• /
• 2008

The sample size in a clinical trial is determined by the hypothesis, the variance of observations, the effect size, the power and the significance level. Dropouts in clinical trials are inevitable, so we need to consider dropouts on the determination of sample size. It is common that some proportion corresponding to the expected dropout rate would be added to the sample size calculated from a mathematical equation. This paper proposes new equations for calculating sample size dealing with dropouts. Since we observe data longitudinally in most clinical trials, we can use a last observation to impute for missing one in the intention to treat (ITT) trials, and this technique is called last observation carried forward(LOCF). But LOCF might make deviations on the assumed variance and effect size, so that we could not guarantee the power of test with the sample size obtained from the existing equation. This study suggests the formulas for sample size involving information about dropouts and shows the properties of the proposed method in testing equality of means.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.18 no.6
• /
• pp.464-476
• /
• 1998

Determination of location, shape and size of a flaw from its eddy current testing signal is one of the fundamental issues in eddy current nondestructive evaluation of steam generator tubes. Here, we propose an approach to this problem; an inversion of eddy current flaw signal using neural networks trained by finite element model-based synthetic signatures. Total 216 eddy current signals from four different types of axisymmetric flaws in tubes are generated by finite element models of which the accuracy is experimentally validated. From each simulated signature, total 24 eddy current features are extracted and among them 13 features are finally selected for flaw characterization. Based on these features, probabilistic neural networks discriminate flaws into four different types according to the location and the shape, and successively back propagation neural networks determine the size parameters of the discriminated flaw.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.2 no.1
• /
• pp.93-104
• /
• 1982

DTM data evaluation regards accuracy and efficiency as its most important parameters, and these two elements are sensitively influenced by the sampling technique employed and its procedure. This study attempts to improve sampling techniques and evaluate the accuracy of DTM based on earth volume calculation, using aerial photography and field survey information as basic DTM data, and a regular grid and progressive sampling method for sampling process. Especially, the progressive sampling has employed different combinations of threshold and sampling criteria, that is, parameters influential on efficiency and accuracy, for the purpose of numerical tests intended to permit comparative analysis. The tests have resulted in the findings that in progressive sampling its grid density fits in well with given terrain characteristics in proportion to their variability. and that threshold and accuracy contain close mutual relations in which an increased threshold bringing more efficency but less accuracy. It follows, therefore. that a threshold, when appropriately determined, can be of very efficient use for considerably broad applications, and particularly, will play a significant role in its application to such general civil engineering as earth volume calculation.

• Korean Journal of Soil Science and Fertilizer
• /
• v.18 no.2
• /
• pp.128-133
• /
• 1985

Variability of soil characteristics and purity were statistically summarized within mapping units and estimates were given for sample sizes required to reach predetermined levels of accuracy in mountainous area soils. Total variances within mapped soil series (expressed as CV): 30-60% for physical properties and CEC relatively affected by management and 80-230% for chemical properties except for pH affected by management. Means of some chemical properties were estimated on the assumption of log-normal distribution. These soils were correctly classified with regard to order at 73.9, to great group 56.9, and to series 47.4%. Approximate sample sizes required to detect the difference of 10 and 20% of the true mean with 0.95 confidence level were presented.

• Korean Journal of Soil Science and Fertilizer
• /
• v.18 no.3
• /
• pp.254-259
• /
• 1985

Variability of soil characteristics and purity were statistically summarized within mapping units and estimates were given for sample sizes required to reach predetermined levels of accuracy in mountain-foot slope soils. Total variances within mapped soil series (expressed as CV): 12-44% for pH, CEC, and particle size distribution was relatively unaffected and 56-196% for OM, ext.-cations, and available $P_2O_5$, affected by management. Means of some chemical properties were estimated on the assumption of lognormal distribution. These soils were correctly classified with regard to order at 52.2%, to great group 53.7%, and to series 39.7%. respectively. Approximate sample sizes required to detect the difference of 10 and 20% of the true mean with 0.95 confidence level were presented.