• The Korean Journal of Applied Statistics
• /
• v.32 no.6
• /
• pp.899-908
• /
• 2019

Massive in-depth neural networks with numerous parameters are powerful machine learning methods, but they have overfitting problems due to the excessive flexibility of the models. Dropout is one methods to overcome the problem of oversized neural networks. It is also an effective method that randomly drops input and hidden nodes from the neural network during training. Every sample is fed to a thinned network from an exponential number of different networks. In this study, instead of feeding one sample for each thinned network, two or more samples are used in fitting for one thinned network known as a Hybrid Dropout. Simulation results using real data show that the new method improves the stability of estimates and reduces the minimum error for the verification data.

• The Journal of the Korea Contents Association
• /
• v.20 no.7
• /
• pp.585-593
• /
• 2020

This study was conducted to research the Selection Attributes of tourists in Jeonju Hanok Village. The purpose of this study was to study the importance and satisfaction after visiting the Jeonju Hanok Village using IPA analysis, and to provide results and marketing implications. The survey was conducted from the October to the November in 2018. A total of 300 questionnaires were distributed and 258 responded questionnaires were reliable to be used as a sample. The result of the survey was analyzed by using SPSS 15.0 version for window with Paired t-test and IPA method. Frequency Analysis was also conducted for the characteristic of samples. As a result of the study, first, the cleanliness of tourist attractions was the highest among the selection attributes, and the next ranking was in the order of parking lot facilities, natural scenery, food, and weather. Second, the natural property was the most satisfactory as a selection property item that tourists visiting Hanok Village were satisfied with, followed by climate(weather), regional characteristics, historical and cultural resources, and cleanliness of tourist attractions. Third, depending on the importance-satisfaction value of the selection attribute variable perceived by tourists visiting Hanok Village, it is necessary to develop various programs in Hanok Village and prepare measures to increase tourist satisfaction.

• The Korean Journal of Financial Management
• /
• v.14 no.2
• /
• pp.273-312
• /
• 1997

본 연구는 우리나라의 실제 일별 주식수익률 자료를 이용한 시뮬레이션을 통해 우리나라 주식시장에 가장 적합한 사건연구방법론을 도출하고자 하였다. 이를 위해, 1980년에서부터 1995년까지 우리나라 주식시장에 상장된 675개 주식을 대상으로 무작위 복원추출 방법에 의해 50개의 개별주식으로 구성된 250개 표본을 선정하였다. 이들 250개 표본을 이용하여 시뮬레이션 기법에 의해 다양한 사건연구 모형의 통계적 오류와 검정력에 미치는 영향을 종합적으로 분석하였다. 시뮬레이션 실험 결과에 의하면, 먼저 사건일을 정확히 포착할 수 있는 경우에는 산업별주가지수를 시장지수로 선택한 시장모형 혹은 시장조정모형으로 초과수익률을 측정하고, 횡단면 독립성을 가정한 검정법이 가장 우수한 사건연구방법으로 나타났다. 한편, 사건일을 정확하게 포착할 수 없는 경우에는, 동일가중지수를 시장지수로 선택한 시장모형 혹은 시장조정모형으로 초과수익률을 측정하고, 횡단면 독립성을 가정한 검정법이 가장 적합한 모형으로 나타났다. 그리고, 사건일 집중현상으로 인한 제1종 오류를 감소시키는 데에는 개별주식의 초과수익률 간의 횡단면 독립성을 가정한 검정법보다는 횡단면 중속성을 조정한 검정법을 사용하는 것이 더욱 효과적이다.

• Journal of Labour Economics
• /
• v.35 no.2
• /
• pp.79-115
• /
• 2012

In the study, we try to get reliable estimates of intergenerational income mobility in Korea. At first, we show that the low estimates of previous studies are mainly due to sample selection problem. The direct estimations using OLS after correcting this problem show higher values than previous estimates. We also compute the attenuation bias by decomposing the variances of earnings into the variances of permanent and transitory components of earnings by the results of the regression. Additionally, we try to estimate the range of intergenerational mobility by comparing the OLS results with the results of the two samples instrumental variable estimation and the three samples instrumental variable estimation. The results of these estimations are a little higher than or similar to OLS results.

• Proceedings of the Korean Society of Computational and Applied Mathematics Conference
• /
• 2003.09a
• /
• pp.2.2-2
• /
• 2003

PRF(Pseudo Random Function)에 대한 랜덤성 검증은 pre-computation공격에 대해 알고리즘이 특별한 통계적 약점이 없이 적절하게 개발되었는지를 평가할 수 있다. 이 논문에서는 NIST에서 실시한 AES 후보 알고리즘 랜덤성 평가 기준을 적용하여 IEEE의 802.11i Draft에서 인증자와 요청자가 비밀키(PTK, GTK)를 생성하는데 사용되는 PRF의 랜덤성을 검증하였다. 랜덤성 테스트를 위해 표본 수는 300개, 표본 길이는 2$^{20}$ (= 1,048,576)으로 검정 표본을 생성하고, 유의 수준은 0.01로 선택하였다. 랜덤성 검증 방법으로는 NIST의 16가지 통계 테스트를 사용하였다.

• The Korean Journal of Applied Statistics
• /
• v.11 no.2
• /
• pp.435-449
• /
• 1998

In this article we have introduced a Bayes criterion for the variable selection in multiple discriminant analysis (MDA). The criterion is a default Bayes factor for the comparision of homo/heteroscadasticity of the multivariate normal means. The default Bayes factor is obtained from a development of the imaginary training sample method introduced by Spiegelhalter and Smith (1982). Based an the criterion, we also provided a test for additional discrimination in MDA. The advantage of the criterion is that it is not only applicable for the optimal subset selection method but for the stepwise method. More over, the criterion can be reduced to that for two-group discriminant analysis. Thus the criterion can be regarded as an unified alternative to variable selection criteria suggested by various sampling theory approaches. To illustrate the performance of the criterion, a numerical study has bean done via Monte Carlo experiment.

• Journal of KIISE:Databases
• /
• v.30 no.5
• /
• pp.464-475
• /
• 2003

few sample objects and compressed histogram information of image databases. The histogram information is used to estimate the selectivity of spherical range queries and a small number of sample objects is used to compensate the selectivity error due to the difference of the similarity measures between meta server and local image databases. An extensive experiment on a large number of image data demonstrates that our proposed method performs well in the distributed heterogeneous environment.

• The Korean Journal of Applied Statistics
• /
• v.5 no.2
• /
• pp.271-282
• /
• 1992

A statistical method to investigate the relationship between the occurrence of Langerahans cells and neoplastic transformation of uterine cerivx. The best fitting submodel which satisfies the selection criterion similar in type to AIC is selected among the possible submodels based on Poisson probability models. A bootstrap method is used to approximate the sampling distribution of the selection criterion and the usual normal approximation is used to find the asymptotic distribution of the estimated rates.

• The Journal of the Korea Contents Association
• /
• v.11 no.11
• /
• pp.472-481
• /
• 2011

The study aims to figure relationships between screen golfers' lifestyle, choice attribute, customer satisfaction, and repurchasing behavior. Study respondents were users of 10 screen golf stores located in Busan and Gyeonnam from March 1, 2010 to April 30, 2010. A total of 290 effective samples were selected based on convenient sampling method and direct visits to stores. SPSS 12.0 was used to conduct factor, reliability, frequent and regression analyses, and the study reached the following conclusions. First, it has been found that screen golfers' lifestyle has a significant effect on choice attribute. Second, it has turn out that screen golfers' choice attribute has a significant effect on customer satisfaction. Third, it has been found that screen golfers' choice attribute has a significant effect on repurchasing behavior.

• The Korean Journal of Applied Statistics
• /
• v.30 no.4
• /
• pp.513-528
• /
• 2017

Statistics are often used to test two samples if they have been drawn from the same underlying distribution. In this paper, we introduce several well-known distribution-free tests to compare distributions and conduct an extensive Monte-Carlo simulation to specify their behaviors. We consider various circumstances of when two distributions vary in (1) location, (2) scale, (3) symmetry, (4) kurtosis, (5) tail weight. A practical guideline for two-sample goodness-of-fit test is presented based on the simulation result.