• Journal of Information Processing Systems
• /
• 제18권2호
• /
• pp.268-281
• /
• 2022

The development of information technology is bringing many changes to everyday life, and machine learning can be used as a technique to solve a wide range of real-world problems. Analysis and utilization of data are essential processes in applying machine learning to real-world problems. As a method of processing data in machine learning, we propose an approach based on applying multiple linear regression models by interlacing data to the task of classifying similar software. Linear regression is widely used in estimation problems to model the relationship between input and output data. In our approach, multiple linear regression models are generated by training on interlaced feature data. A combination of these multiple models is then used as the prediction model for classifying similar software. Experiments are performed to evaluate the proposed approach as compared to conventional linear regression, and the experimental results show that the proposed method classifies similar software more accurately than the conventional model. We anticipate the proposed approach to be applied to various kinds of classification problems to improve the accuracy of conventional linear regression.

• KIEE International Transactions on Electrophysics and Applications
• /
• 제3C권1호
• /
• pp.1-4
• /
• 2003

This paper presents an attempt to develop probabilistic multistress life models to evaluate the lifetime characteristics of epoxy-encapsulated magnet wire with heavy build polyurethane enamel. A set of accelerated life tests were conducted over a wide range of pulsating voltages, temperatures, and frequencies. Samples of fine gauge twisted pairs of the encapsulated magnet wire were tested us-ing a pulse endurance dielectric test system. An electrical-thermal lifetime function was combined with the Weibull distribution of lifetimes. The parameters of the combined Weibull-electrical-thermal model were estimated using maximum likelihood estimation. Likewise, a generalized electrical-thermal-frequency life model was also developed. The parameters of this new model were estimated using multiple linear regression technique. It was found in this paper that lifetime estimates of the two proposed probabilistic multistress life models are good enough. This suggests the suitability of using the general electrical-thermal-frequency model to estimate the lifetime of the encapsulated magnet wire over a wide range of voltages, temperatures and pulsating frequencies.

• 응용통계연구
• /
• 제26권3호
• /
• pp.531-547
• /
• 2013

연생모형(multiple life model)은 보험계약에서 두 명 또는 그 이상의 피보험자들의 사망 또는 생존의 상태에 따른 보험금을 지급하는 보험 상품의 보험료 결정 및 리스크 관리를 위한 모형이다. 본 논문에서는 부부 가입자 중 마지막 사망자가 발생할 때까지 연금이 지급되는 연생보험의 대표적인 상품인 역모기지를 살펴보고자 한다. 역모기지 상품은 만 60세 이상의 고령자가 소유주택을 담보로 거주하면서 일정액을 대출받지만 대출금을 다시 상환하지 않는 금융상품으로 부부 가입의 경우 대출 기간의 확률분포를 적용하기 위해서는 연생모형을 적용해야만 한다. 그러나 현행 역모기지 상품에 있어서 대출한도 및 월지급금 산출 시 연생모형이 적용되지 않고 있으며 우리나라의 경우에는 국민 생명표 상의 여자 사망률을 대출 종료(연금지급종료)확률로 활용하고 있다. 여자의 사망률을 이용하는 이유는 보수적인 관점에서 대출 종료 시점을 예측하기 위해 일반적으로 남자보다 여자가 더 수명이 길다는 점 때문이다. 고령화로 인해 수명이 점점 길어지는 추세이기 때문에 역모기지와 같이 계약기간이 확정되어 있지 않은 보험 상품의 경우 특히 더 계약 종료 시점에 대한 확률분포가 리스크 관리를 위하여 중요하다. 본 논문의 의의는 역모기지의 발행기관 및 보증기관의 적정한 월지급금 지급과 차후 월지급금의 과대지급으로 인한 지급불능을 방지하기 위해 현행 사용하고 있는 모형의 위험률에서 연생 모형으로 변경할 필요성을 실증분석을 통하여 제시한다.

• Asian-Australasian Journal of Animal Sciences
• /
• 제15권7호
• /
• pp.925-931
• /
• 2002

Gibbs sampling algorithms were implemented to the multi-trait threshold animal models with any combinations of multiple binary, ordered categorical, and linear traits and investigate the amount of bias on these models with two kinds of parameterization and algorithms for generating underlying liabilities. Statistical models which included additive genetic and residual effects as random and contemporary group effects as fixed were considered on the models using simulated data. The fully conditional posterior means of heritabilities and genetic (residual) correlations were calculated from 1,000 samples retained every 10th samples after 15,000 samples discarded as "burn-in" period. Under the models considered, several combinations of three traits with binary, multiple ordered categories, and continuous were analyzed. Five replicates were carried out. Estimates for heritabilities and genetic (residual) correlations as the posterior means were unbiased when underlying liabilities for a categorical trait were generated given by underlying liabilities of the other traits and threshold estimates were rescaled. Otherwise, when parameterizing threshold of zero and residual variance of one for binary traits, heritability estimates were inflated 7-10% upward. Genetic correlation estimates were biased upward if positively correlated and downward if negatively correlated when underling liabilities were generated without accounting for correlated traits on prior information. Residual correlation estimates were, consequently, much biased downward if positively correlated and upward if negatively correlated in that case. The more categorical trait had categories, the better mixing rate was shown.

• 한국초지조사료학회지
• /
• 제36권3호
• /
• pp.223-236
• /
• 2016

The objective of this study was to construct Italian ryegrass (IRG) dry matter yield (DMY) estimation models in South Korea based on climatic data by locations. Obviously, the climatic environment of Jeju Island has great differences with Korean Peninsula. Meanwhile, many data points were from Jeju Island in the prepared data set. Statistically significant differences in both DMY values and climatic variables were observed between south areas of Korean Peninsula and Jeju Island. Therefore, the estimation models were constructed separately for south areas of Korean Peninsula and Jeju Island separately. For south areas of Korean Peninsula, a data set with a sample size of 933 during 26 years was used. Four optimal climatic variables were selected through a stepwise approach of multiple regression analysis with DMY as the response variable. Subsequently, via general linear model, the final model including the selected four climatic variables and cultivated locations as dummy variables was constructed. The model could explain 37.7% of the variations in DMY of IRG in south areas of Korean Peninsula. For Jeju Island, a data set containing 130 data points during 17 years were used in the modeling construction via the stepwise approach of multiple regression analysis. The model constructed in this research could explain 51.0% of the variations in DMY of IRG. For the two models, homoscedasticity and the assumption that the mean of the residuals were equal to zero were satisfied. Meanwhile, the fitness of both models was good based on most scatters of predicted DMY values fell within the 95% confidence interval.

• 국제학술발표논문집
• /
• The 9th International Conference on Construction Engineering and Project Management
• /
• pp.551-562
• /
• 2022

The building production system has been analysed by the dichotomy "employer-contractor" relationship, which failed to take into account of the role and function of multiple stakeholders within the life-cycle supply chain. This is further observed in the current conflict resolution model, which, in my argument, struggles to contribute to industrialize the building production and achieve better efficiency and effectiveness as expected. The purpose of this paper is to critically assess the issues of current programme-based conflict resolution model, and discuss alternative models how they can be modelled and applied to the construction projects. The conclusions of findings are; First, the current model is framed around the contracts and dispute resolutions based on the legal concept of "claimant and respondent" where one party(s) advances a claim once and the other(s) objects, as such it fails to reflect the nature of construction projects where multiple stakeholders are involved concurrently and for a long period of life-cycle of buildings. Second, an alternative is "Six-stakeholders model" which represents the multiple stakeholders and clarifies the flow of obligation-liability-monetary relationships among participants for a long period of life-cycle of buildings. Further, with reference to both historical and recent cases, a reflection and insight into pros and cons of programming method is added, especially as to why this method is considered to have become a mandate of the modern construction management, and how academics and practitioners should deal with it more cautiously and prudently.

• 대한간호학회지
• /
• 제19권2호
• /
• pp.173-190
• /
• 1989

This literature review was undertaken to explore theoretical models of depression for their potential usefulness in nursing research and practice. Depression has bean accounted for by numerous theories or models of causation ; 11 theories selected from psychology, medicine and psychoanalysis and supported by empirical or experimental research were reviewed. These theories identify a variety of precipitating and predisposing factors that may affect the individual's depression. Aggression - turned - inward theory, object loss theory, ego functioning theory, personality organization theory, behavioral theory, learned helplessness theory, cognitive theory, genetic factors, and biological theories conceptualize predisposing factors. Only life stressors theory identifies precipitating facotrs. Each of these theories contributes to an understanding of depression, but many of them use overlapping and interrelated factors. It is also evident from recent. research that there are multiple causes for depression involving an interactive effect among predisposing and precipitating factors that are both biological and psychological in origin. That is, a single theory is not useful, but perhaps a unified theory could be developed that would be helpful to nursing. This review points to the need for continuing development and testing of theories that would integrate the multiple conceptualizations of depression.

• Molecules and Cells
• /
• 제42권3호
• /
• pp.237-244
• /
• 2019

Understanding the mechanisms of cancer drug resistance is a critical challenge in cancer therapy. For many cancer drugs, various resistance mechanisms have been identified such as target alteration, alternative signaling pathways, epithelial-mesenchymal transition, and epigenetic modulation. Resistance may arise via multiple mechanisms even for a single drug, making it necessary to investigate multiple independent models for comprehensive understanding and therapeutic application. In particular, we hypothesize that different resistance processes result in distinct gene expression changes. Here, we present a web-based database, CDRgator (Cancer Drug Resistance navigator) for comparative analysis of gene expression signatures of cancer drug resistance. Resistance signatures were extracted from two different types of datasets. First, resistance signatures were extracted from transcriptomic profiles of cancer cells or patient samples and their resistance-induced counterparts for >30 cancer drugs. Second, drug resistance group signatures were also extracted from two large-scale drug sensitivity datasets representing ~1,000 cancer cell lines. All the datasets are available for download, and are conveniently accessible based on drug class and cancer type, along with analytic features such as clustering analysis, multidimensional scaling, and pathway analysis. CDRgator allows meta-analysis of independent resistance models for more comprehensive understanding of drug-resistance mechanisms that is difficult to accomplish with individual datasets alone (database URL: http://cdrgator.ewha.ac.kr).

• 한국균학회소식:학술대회논문집
• /
• 한국균학회 2018년도 춘계학술대회 및 임시총회
• /
• pp.22-22
• /
• 2018

Fungal genome sequencing and assembly have been trivial in these days. Genome analysis relies on high quality of gene prediction and annotation. Automatic fungal genome annotation pipeline is essential for handling genomic sequence data accumulated exponentially. However, building an automatic annotation procedure for fungal genomes is not an easy task. FunGAP (Fungal Genome Annotation Pipeline) is developed for precise and accurate prediction of gene models from any fungal genome assembly. To make high-quality gene models, this pipeline employs multiple gene prediction programs encompassing ab initio, evidence-, and homology-based evaluation. FunGAP aims to evaluate all predicted genes by filtering gene models. To make a successful filtering guide for removal of false-positive genes, we used a scoring function that seeks for a consensus by estimating each gene model based on homology to the known proteins or domains. FunGAP is freely available for non-commercial users at the GitHub site (https://github.com/CompSynBioLab-KoreaUniv/FunGAP).

• Journal of Animal Science and Technology
• /
• 제66권1호
• /
• pp.167-177
• /
• 2024

Iris segmentation is an initial step for identifying the biometrics of animals when establishing a traceability system for livestock. In this study, we propose a deep learning framework for pixel-wise segmentation of bovine iris with a minimized use of annotation labels utilizing the BovineAAEyes80 public dataset. The proposed image segmentation framework encompasses data collection, data preparation, data augmentation selection, training of 15 deep neural network (DNN) models with varying encoder backbones and segmentation decoder DNNs, and evaluation of the models using multiple metrics and graphical segmentation results. This framework aims to provide comprehensive and in-depth information on each model's training and testing outcomes to optimize bovine iris segmentation performance. In the experiment, U-Net with a VGG16 backbone was identified as the optimal combination of encoder and decoder models for the dataset, achieving an accuracy and dice coefficient score of 99.50% and 98.35%, respectively. Notably, the selected model accurately segmented even corrupted images without proper annotation data. This study contributes to the advancement of iris segmentation and the establishment of a reliable DNN training framework.