• Journal of the Korean Statistical Society
• /
• 제27권4호
• /
• pp.459-469
• /
• 1998

Balanced half sample method is a simple variance estimation method for complex sampling designs. Since it is simple and flexible, it has been widely used in large scale sample surveys. However, the usual BHS method overestimate the true variance in without replacement sampling and two-stage cluster sampling. Focusing on this point , we proposed an unbiased BHS variance estimator in a stratified two-stage cluster sampling and then described an implementation method of the proposed estimator. Finally, partially BHS design is explained as a tool of reducing the number of replications of the proposed estimator.

• 품질경영학회지
• /
• 제26권4호
• /
• pp.239-249
• /
• 1998

In balanced design of experiment sample size is determined to detect a certain standardized size of effect with power 1-$\beta$ at the level of significance $\alpha$. Tables (Marvin, et al (1970) and Lorenzen and Anderson(1993) and charts (Odeh and Fox(1991)) are available to determine the sample size in balanced design of experiments. To simplify those tables and charts simple MATLAB program is used to find the minimal standardized detectable difference $\delta$ when $\alpha , \beta$ and sample size are given.

• Journal of the Korean Statistical Society
• /
• 제32권3호
• /
• pp.245-259
• /
• 2003

The 2-way balanced one-level rotation design has been discussed (Park et al., 2001), where the 2-way balancing is done on interview time in monthly sample and rotation group. We extend it to 3-way balanced multi-level design to obtain more information of the same sample unit for one or more previous months. The 3-way balancing is accomplished not only on interview time in monthly sample and rotation group but also on recall time as well. The 3-way balancing eliminates or reduces any bias arising from unbalanced interview time, rotation group and recall time, and all rotation groups are equally represented in the monthly sample. We present the rule and rotation algorithm which guarantee the 3-way balancing. In particular, we specify the necessary and sufficient condition for the 3-way balanced multi-level rotation design.

• Communications for Statistical Applications and Methods
• /
• 제14권3호
• /
• pp.667-677
• /
• 2007

In this paper, we propose a method for estimating the mean of a population which has a linear trend, when both n, the sample size, and k, the reciprocal of the sampling fraction, are odd numbers. The proposed method, not having the drawbacks of centered systematic sampling, centered modified sampling and centered balanced sampling, consists of selecting a sample by balanced systematic sampling and estimating the population mean by using interpolation. We compare the efficiency of the proposed method and existing methods under the criterion of the expected mean square error based on the infinite superpopulation model.

• 한국통계학회:학술대회논문집
• /
• 한국통계학회 2002년도 춘계 학술발표회 논문집
• /
• pp.19-24
• /
• 2002

The two-way balanced one-level rotation design has been discussed (Park, Kim and Choi, 2001), where the two-way balancing is done on interview time in monthly sample and rotation group. We extend it to three-way balanced multi-level design under the most general rotation system. The three-way balancing is accomplished on interview time not only in monthly sample and rotation group but also in recall time. We present the necessary condition and rotation algorithm which guarantee the three-way balancing. We propose multi-level composite estimators (MCE) from this design and derive their variances and mean squared errors (MSE), assuming the correlation from the measurements of the same sample unit and three types of biases in monthly sample.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송∙미디어공학회 2022년도 추계학술대회
• /
• pp.164-166
• /
• 2022

Long-tail problem은 class 별로 sample의 개수에 차이가 있어 성능에 안 좋은 영향을 미치는 것을 말한다. 본 논문에서는 cost-sensitive learning 중 Class-Balanced Loss를 이용해 성능을 개선하여 Long-tail problem을 해결하려고 한다. 먼저, balanced data set과 imbalanced data set의 성능 차이를 살펴보도록 할 것이다. 그 후, Class-Balanced Loss를 3가지 버전으로 이용해 그 성능을 측정하고 분석해 볼 것이다.

• Communications for Statistical Applications and Methods
• /
• 제9권3호
• /
• pp.889-898
• /
• 2002

In decision theoretic estimation, the loss function usually emphasizes precision of estimation. However, one may have interest in goodness of fit of the overall model as well as precision of estimation. From this viewpoint, Zellner(1994) proposed the balanced loss function which takes account of both "goodness of fit" and "precision of estimation". This paper considers estimation of the parameter of a Bernoulli distribution using Zellner's(1994) balanced loss function. It is shown that the sample mean $\overline{X}$, is admissible. More general results, concerning the admissibility of estimators of the form $a\overline{X}+b$ are also presented. Finally, minimax estimators and some numerical results are given at the end of paper,at the end of paper.

• Journal of Information Processing Systems
• /
• 제10권2호
• /
• pp.314-333
• /
• 2014

When analyzing default predictions in real estate companies, the number of non-defaulted cases always greatly exceeds the defaulted ones, which creates the two-class imbalance problem. This lowers the ability of prediction models to distinguish the default sample. In order to avoid this sample selection bias and to improve the prediction model, this paper applies a minority sample generation approach to create new minority samples. The logistic regression, support vector machine (SVM) classification, and neural network (NN) classification use an imbalanced dataset. They were used as benchmarks with a single prediction model that used a balanced dataset corrected by the minority samples generation approach. Instead of using prediction-oriented tests and the overall accuracy, the true positive rate (TPR), the true negative rate (TNR), G-mean, and F-score are used to measure the performance of default prediction models for imbalanced dataset. In this paper, we describe an empirical experiment that used a sampling of 14 default and 315 non-default listed real estate companies in China and report that most results using single prediction models with a balanced dataset generated better results than an imbalanced dataset.

• Journal of Computing Science and Engineering
• /
• 제10권4호
• /
• pp.118-127
• /
• 2016

An improved sample balanced genetic algorithm and Extreme Learning Machine (iSBGA-ELM) was designed for accurate diagnosis of Alzheimer disease (AD) and identification of biomarkers associated with AD in this paper. The proposed AD diagnosis approach uses a set of magnetic resonance imaging scans in Open Access Series of Imaging Studies (OASIS) public database to build an efficient AD classifier. The approach contains two steps: "voxels selection" based on an iSBGA and "AD classification" based on the ELM. In the first step, the proposed iSBGA searches for a robust subset of voxels with promising properties for further AD diagnosis. The robust subset of voxels chosen by iSBGA is then used to build an AD classifier based on the ELM. A robust subset of voxels keeps a high generalization performance of AD classification in various scenarios and highlights the importance of the chosen voxels for AD research. The AD classifier with maximum classification accuracy is created using an optimal subset of robust voxels. It represents the final AD diagnosis approach. Experiments with the proposed iSBGA-ELM using OASIS data set showed an average testing accuracy of 87%. Experiments clearly indicated the proposed iSBGA-ELM was efficient for AD diagnosis. It showed improvements over existing techniques.

• 한국생활과학회지
• /
• 제13권5호
• /
• pp.691-699
• /
• 2004

The purpose of this study was to analyze the relationship between family functioning and family violence. The SPSS 10.0 for Windows was used to analyze data obtained through 1,044 children who live in Gwangju Chonnam area. Major findings are as follows: 1. Wife abuse and child abuse showed significant difference according to family cohesion, family flexibility, and family-system types. 2. In family violence non-experience group, 'balanced family' was found in 36.2% of the sample and 'extreme family' in 7.8%. In family violence experience group, 'balanced family' was found in 12.7% of the sample and 'extreme family' in 15.4%. 3. There were significant negative correlations between family functioning and family violence. Family cohesion and family flexibility had significant negative correlation with wife abuse and child abuse.