• The Korean Journal of Applied Statistics
• /
• v.17 no.3
• /
• pp.445-457
• /
• 2004

Given class-imbalanced data in two-class classification problem, we often do over-sampling and/or under-sampling of training data to make it balanced. We investigate the validity of such practice. Also we study the effect of such sampling practice on boosting of classification trees. Through experiments on twelve real datasets it is observed that keeping the natural distribution of training data is the best way if you plan to apply boosting methods to class-imbalanced data.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2014.01a
• /
• pp.145-147
• /
• 2014

공동주의는 사회적 개체간의 지각적 경험을 공유하는 상호작용과정으로, 최근 인간-로봇 상호작용연구와 관련해서 로봇 공학자의 관심이 커지고 있다. 발달심리학에 기초한 기존의 developmental robotics의 접근과는 달리, 본 연구에서는 사전단서 패러다임을 이용해서 목표물, 단서, 주의반응 간의 삼자관계를 수학적으로 개념화하였다. 간단한 목표물 탐사과제를 통해서 계산모형의 수행을 검증하였다. 연구결과에서는 컴퓨터 시스템의 시각적 주의 모형이 사용자가 지시하는 단서(손가락 지시)의해 목표물(이온음료)을 주의를 할당하는 것을 보였다. 본 연구는 심리학에서 연구된 사전단서 패러다임을 인간-로봇 상호작용에 적용될 수 있음을 보여준다.

• The Korean Journal of Applied Statistics
• /
• v.28 no.2
• /
• pp.269-279
• /
• 2015

This paper considers inferences of random effects. We show that the proposed confidence distribution (CD) performs well in logistic regression for random intercepts with small samples. Real data analyses are also done to identify the subject effects clearly.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2012.05a
• /
• pp.665-665
• /
• 2012

식생의 물리적 특성에 따라 조도의 관점에서 식생이 흐름에 미치는 영향도 많이 달라지게 된다. 일반적인 식생 조도 실험은 초본류를 대상으로 하고 있지만, 초본류 식생은 매우 유연하기 때문에 단일 개채가 흐름에 미치는 영향을 정량적으로 파악하는 것이 매우 힘들다. 따라서 단일 개체에 의한 조도 특성 보다는 식생군에 의한 조도계수 변화 특성을 알아내는 것이 보통 중요해진다. 하지만 조도의 관점에서 초본류와 달리 목본류는 삭생의 강성이 매우 크기 때문에 식생군의 군집으로서 특징보다는 개별 식생의 형상이 큰 역할을 수행하게 된다. 일반적으로 목본류는 수심에 따라 흐름에 영향을 주는 생체량(biosmass)이 수심에 따라 변화하지만, 줄기(trunk)의 경우 수심에 따라 생체량의 변화가 거의 없으며, 줄기의 강성 및 단면 모양이 큰 역할을 하게 된다. 또한 목본류는 실험 수로에 고밀도의 수목군을 형성하여 실험을 수행하는 것이 쉽지 않으므로 개별 식생이나 소규모 군체에 대하여 항력을 측정하여 조도를 평가하는 것이 일반적이다. 따라서 본 연구에서는 목본류의 줄기 부분이 흐름에 미치는 영향을 파악하기 위한 기초 연구로서 원기둥이나 사각기둥과 같은 식생 모형 요소를 이용하여 식생 줄기를 구현하였고, 식생 형상 및 식생 직경 등을 달리하여 침수 조건 변화에 따른 항력 변화 특성을 검토하였다.

• Journal of Animal Science and Technology
• /
• v.48 no.1
• /
• pp.29-38
• /
• 2006

The objective of this study was to estimate genetic variances of growth curve parameters in Hanwoo cows. The data used in this study were records from 1,083 Hanwoo cows raised at Hanwoo Experiment Station, National Livestock Research Institute(NLRI). First evaluation model(Model I) fit year-season of birth and age of dam as fixed effects and second model(Model II) added age at the final weight as a linear covariate to Model I. Heritability estimates of A, b and k from Gompertz model were 0.22, 0.11 and 0.07 using modelⅠ and 0.28, 0.11 and 0.12 using modelⅡ. Those from Von Bertalanffy model were 0.22, 0.11 and 0.07 using modelⅠ, 0.28, 0.11 and 0.12 using modelⅡ. Heritability estimates of A, b and k from Logistic model were 0.14, 0.07 and 0.05 using modelⅠ, 0.18, 0.07 and 0.12 using modelⅡ. Heritability estimates of A from Gompertz model were higher than those from Von Bertalanffy model or Logistic model in both model Ⅰand model Ⅱ. Heritability estimates of b from Logistic model were higher than those from Gompertz model or Von Bertalanffy model in both modelⅠand model Ⅱ. Heritability estimates of birth weight, weaning weight, 3 month weight, 6 month weight, 9 month weight, 12 month weight, 18 month weight, 24 month weight, 36 month weight were after linear age adjustment 0.27, 0.11, 0.19, 0.14, 0.16, 0.23, 0.52 and 0.32, respectively. Heritability estimates of birth weight, weaning weight, 3 month weight, 6 month weight, 9 month weight and 24 month weight fit by Gompertz model were larger than those estimated from linearly adjusted data. Heritability estimates of 12 month weight, 18 month weight and 36 month weight fit by Von Bertalanffy model were larger than those estimated from linearly adjusted data. In the multitrait analyses for parameters from Gompertz model, genetic and phenotypic correlations between A and k parameters were －0.47 and －0.67 using modelⅠand －0.56 and －0.63 using model Ⅱ. Those between the A and b parameters were 0.69 and 0.34 using modelⅠand 0.72 and 0.37 using model Ⅱ. Those between the b and k parameters were －0.26 and 0.01 using modelⅠand －0.30 and 0.01 using model Ⅱ. In the multitrait analyses for parameters from Von Bertalanffy model, genetic and phenotypic correlations between A and k parameters were －0.49 and －0.67 suing model Ⅰ and －0.57 and －0.70 using modelⅡ. Those between the A and b parameters were 0.61 and 0.33 using modelⅠ and 0.60 and 0.30 using model Ⅱ. Those between the b and k parameters were －0.20 and 0.02 using modelⅠ and 0.16 and 0.00 using modelⅡ. In the multitrait analyses for parameters from Logistic model, genetic and phenotypic correlations between A and k parameters were －0.43 and －0.67 using model Ⅰ and －0.50 and －0.63 using modelⅡ. Those between the A and b parameters were 0.47 and 0.22 using modelⅠ and 0.38 and 0.24 using modelⅡ. Those between the b and k parameters were －0.09 and 0.02 using model Ⅰ and －0.02 and 0.13 using model Ⅱ.

• The Korean Journal of Applied Statistics
• /
• v.25 no.2
• /
• pp.279-291
• /
• 2012

When fitting a Cox proportional hazards model with missing covariates, it is inefficient to exclude observations with missing values in the analysis. Furthermore, if the missing-data mechanism is not Missing Completely At Random(MCAR), it may lead to biased parameter estimation. Many approaches have been suggested to handle the Cox proportional hazards model when covariates are sometimes missing, but they are based on the selection model. This paper suggest an approach to handle Cox proportional hazards model with missing covariates by using the pattern-mixture model (Little, 1993). The pattern-mixture model is expressed by the joint distribution of survival time and the missing-data mechanism. In the pattern-mixture model, many models can be considered by setting up various restrictions, and different results under various restrictions indicate the sensitivity of the model due to missing covariates. A simulation study was conducted to show the sensitivity of parameter estimation under different restrictions in a pattern-mixture model. The proposed approach was also applied to mouse leukemia data.

• The KIPS Transactions:PartB
• /
• v.11B no.4
• /
• pp.485-490
• /
• 2004

In data representation, the clustering performs a grouping process which combines given data into some similar clusters. The various similarity measures have been used in many researches. But, the validity of clustering results is subjective and ambiguous, because of difficulty and shortage about objective criterion of clustering. The fuzzy clustering provides a good method for subjective clustering problems. It performs clustering through the similarity matrix which has fuzzy membership value for assigning each object. In this paper, for objective fuzzy clustering, the clustering algorithm which joins principal components analysis as a dimension reduction model with bayesian learning as a statistical learning theory. For performance evaluation of proposed algorithm, Iris and Glass identification data from UCI Machine Learning repository are used. The experimental results shows a happy outcome of proposed model.

• Environmental and Resource Economics Review
• /
• v.20 no.2
• /
• pp.357-379
• /
• 2011

The problem of invasive species has been recently emerged as one of complicated issues due to increasing globalisation and its consequence of species immigrations. Since in most cases of invasive species it is less likely to fully eradicate them through human efforts, it is often interested in reducing the possibility of ecological disaster caused by the invasive species. This paper provides an optimal control model to minimize such possibility while allowing the stochastic nature of biological growth of the invasive species. Conditions under which the partial eradication effort is optimal are derived. Simple numerical illustration is provided using H1N1 data which is categorized as an invasive disease in microorganism level.

• Journal of Animal Science and Technology
• /
• v.55 no.5
• /
• pp.359-365
• /
• 2013

Estimate breeding value can be used as single trait animal model was developed directly using the Fortran language program. The program is based on data computed by using the indirect method repeatedly. The program develops a common algorithm and imprves efficiency. Algorithm efficiency was compared between the two programs. Estimated using the solution is easy to farm and brand the service, pedigree data base was associated with the development of an improved system. The existing program that uses the single trait animal model and the comparative analysis of efficiency is weak because the estimation of the solution and the conventional algorithm programmed through regular formulation involve many repetition; therefore, the newly developed algorithm was conducted to improve speed by reducing the repetition. Single trait animal model was used to analyze Gauss-Seidel iteration method, and the aforesaid two algorithms were compared thorough the mixed model equation which is used the most commonly in estimating the current breeding value by applying the procedures such as the preparation of information necessary for modelling, removal of duplicative data, verifying the parent information of based population in the pedigree data, and assigning sequential numbers, etc. The existing conventional algorithm is the method for reading and recording the data by utilizing the successive repetitive sentences, while new algorithm is the method for directly generating the left hand side for estimation based on effect. Two programs were developed to ensure the accurate evaluation. BLUPF90 and MTDFREML were compared using the estimated solution. In relation to the pearson and spearman correlation, the estimated breeding value correlation coefficients were highest among all traits over 99.5%. Depending on the breeding value of the high correlation in Model I and Model II, accurate evaluation can be found. The number of iteration to convergence was 2,568 in Model I and 1,038 in Model II. The speed of solving was 256.008 seconds in Model I and 235.729 seconds in Model II. Model II had a speed of approximately 10% more than Model I. Therefore, it is considered to be much more effective to analyze large data through the improved algorithm than the existing method. If the corresponding program is systemized and utilized for the consulting of farm and industrial services, it would make contribution to the early selection of individual, shorten the generation, and cultivation of superior groups, and help develop the Hanwoo industry further through the improvement of breeding value based enhancement, ultimately paving the way for the country to evolve into an advanced livestock country.

• The Korean Journal of Malacology
• /
• v.24 no.2
• /
• pp.121-130
• /
• 2008

In other aquaculture species, large improvements in growth have been achieved through selective breeding. Ezo abalone(Haliotis discus hannai) and disk abalone(H. discus discus) are major aquatic animals cultured in Asia, but selective breeding for the promotion of growth with these abalones has not been actively pursued. Recently significant efforts are being made to promote production of these species through selective breeding in Korea. The aims of this work were to estimate the general genetic parameters, heritabilities, and genetic and phenotypic correlations on growth-related traits at 1-year old in two Korean abalone subspecies, H. discus hannai and H. discus discus, by using multiple trait animal model. The data were collected from the records of 1,504 individuals produced from 22 sires and 26 dams in H. discus hannai and 297 individuals produced from 5 sires and 6 dams in H. discus discus, which evaluated by the Genetics and Breeding Research Center, National Fisheries Research & Development Institute(NFRDI). Genetic parameters were estimated for these abalone subspecies raised in Bukjeju branch, NFRDI, from May 20, 2004 to May 16, 2005, respectively. The heritability estimates obtained from restricted maximum likelihood(REML) were higher than expected, ranging from 0.40 to 0.43 for growth traits shell length, shell width and body weight in H. discus hannai and from 0.26 to 0.51 in H. discus discus, respectively. The heritabilities for shell shape and condition factor were lower than others of growth traits such as ranging from 0.09 to 0.19 in H. discus hannai and from 0.10 to 0.23 in H. discus discus, respectively. Genetic and phenotypic were > 0.93 between shell parameters and weight in two abalone species, respectively, indicating that breeding for weight gains could be successfully achieved by selecting for shell length.