• Industrial Engineering and Management Systems
• /
• 제14권2호
• /
• pp.210-219
• /
• 2015

A classification task requires an exponentially growing amount of computation time and number of observations as the variable dimensionality increases. Thus, reducing the dimensionality of the data is essential when the number of observations is limited. Often, dimensionality reduction or feature selection leads to better classification performance than using the whole number of features. In this paper, we study the possibility of utilizing the Markov blanket discovery algorithm as a new feature selection method. The Markov blanket of a target variable is the minimal variable set for explaining the target variable on the basis of conditional independence of all the variables to be connected in a Bayesian network. We apply several Markov blanket discovery algorithms to some high-dimensional categorical and continuous data sets, and compare their classification performance with other feature selection methods using well-known classifiers.

• 한국가스학회지
• /
• 제22권4호
• /
• pp.13-26
• /
• 2018

현재 KS Code 등 국내규정에서는 폭발위험장소의 범위를 계산하는 방법이 명확하게 나타나지 않아, 정확한 범위 선정을 위해서는 확산 모델링 해석을 이용하여야 한다. 본 연구애서는 대표적인 물질과 운전조건을 활용하여 확산 모델링에 비하여 간편하면서도 비교적 합리적인 폭발위험장소의 범위를 산정하는 방법을 제시하고자 하였다. 현재 시행되고 있는 국내외 표준을 바탕으로 폭발하한계(LFL, Lower Flammable Limit)까지 거리에 영향을 미치는 변수를 선정하였다. 총 16종의 인화성물질을 대상으로 물질변수, 운전변수, 기상조건에 대하여 모델링을 진행하였으며, 통계분석을 통해 영향을 미치는 변수를 선별하였다. 선별된 변수를 이용하여 폭발위험장소의 범위 선정을 위한 3단계 분류화 방법(3Step Classification Method)을 작성하였다.

• 사상체질의학회지
• /
• 제19권1호
• /
• pp.98-115
• /
• 2007

1. Objectives This study focuses on the Validity of the Questionnaire about Sasang Constitution Classification for Mongolians 2. Methods By using the way of backward elimination, certain variables are chosen from the 438 cases whose physical conditions are absolutely diagnosed. After that, discriminant analysis for the selected variables has been done to obtain the physical constitution equation and the accuracy ratio of diagnosis which are useful for physical constitution diagnosis. 3. Results and Conclusions (1) In tile Validity for the Questionnaire of Sasang Constitution Classification for Mongolians, the accuracy ratio of diagnosis of Taeyangin is 100%, Soyangin 62.5%, Taeumin 76.7%, and Soeumin 66.1% respectively as a result of the discriminant analysis employing Cronbach's alpha coefficient. On the whole, the accuracy ratio of diagnosis is 70.1%. (2). In the Validity for the Questionnaire of Sasang Constitution Classification for Mongolians, the accuracy ratio of diagnosis of 70.1% means that it beats the maximum chance criterion of 41.4% and the proportional chance criterion of 34.4% by 28.7% and 35.7% respectively. Conclusively, this questionnaire has discriminant power.

• 한국소음진동공학회논문집
• /
• 제21권11호
• /
• pp.1020-1028
• /
• 2011

As rotating machines play an important role in industrial applications such as aeronautical, naval and automotive industries, many researchers have developed various condition monitoring system and fault diagnosis system by applying artificial neural network. Since using obtained signals without preprocessing as inputs of neural network can decrease performance of fault classification, it is very important to extract significant features of captured signals and to apply suitable features into diagnosis system according to the kinds of obtained signals. Therefore, this paper proposes a neural-network-based fault diagnosis system using AR coefficients as feature vectors by LPC(linear predictive coding) and EIV(errors-in variables) analysis. We extracted feature vectors from sound, vibration and current faulty signals and evaluated the suitability of feature vectors depending on the classification results and training error rates by changing AR order and adding noise. From experimental results, we conclude that classification results using feature vectors by EIV analysis indicate more than 90 % stably for less than 10 orders and noise effect comparing to LPC.

• 한국빅데이터학회지
• /
• 제4권2호
• /
• pp.61-71
• /
• 2019

반도체 웨이퍼의 표면을 연마하여 평탄화하는 Chemical Mechanical Planarization(CMP) 공정은 다양한 화학물질과 물리적인 기계장치에 의한 작용을 받기 때문에 공정을 안정적으로 관리하기 힘들다. CMP 공정에서 품질 지표로는 Material Removal Rate(MRR)를 많이 사용하고, CMP 공정의 안정적 관리를 위해서는 MRR을 예측하는 것이 중요하다. 본 연구에서는 머신러닝 기법들을 이용하여 CMP 공정에서 수집된 시계열 센서 데이터를 분석하여 MRR을 예측하는 모형과 공정 품질을 해석하기 위한 분류 모형을 개발한다. 나아가 분류 결과를 분석하여, CMP 공정 품질에 영향을 미치는 유의미한 변수를 파악하고 고품질을 유지하기 위한 공정 조건을 설명한다.

• 응용통계연구
• /
• 제30권6호
• /
• pp.851-865
• /
• 2017

최근 워너크라이라는 이름의 랜섬웨어가 전 세계적으로 큰 화두에 오르면서, 악성 소프트웨어로 인한 피해를 줄이기 위한 방법들이 재조명 되고 있다. 새로운 악성 소프트웨어가 발생했을 때 피해를 최소화하기 위해서는 해당 소프트웨어가 어떤 공격 유형을 가진 악성 소프트웨어인지 빠르게 분류할 필요가 있다. 본 연구 목적은 다양한 통계적 기법을 이용하여 악성 소프트웨어를 효과적으로 분류할 수 있는 모형을 구축하는 데 있다. 모형 적합 시 다항 로지스틱, 랜덤 포레스트, 그래디언트 부스팅, 서포트 벡터 기계 등의 기법들을 이용하였으며, 본 연구를 통해 악성 소프트웨어를 분류하는 데에 있어 중요한 역할을 하는 변수들이 존재한다는 사실을 발견하였다.

• 한국의류학회지
• /
• 제40권2호
• /
• pp.287-304
• /
• 2016

This study examines the influence of product classification standards and structure on user perception as well as their attitude towards online shopping sites. The causal relationships of variables are also examined. The analysis was based on an online survey with 247 responses. Four types of internet shopping sites were developed and used as a stimulus. The results of the mean comparison analysis indicated that perceived variety, information overload, perceived shopping value and attitude towards the site varies significantly with product classification standards and structure. There was also of a marginally significant interaction between the classification standard and structure on perceived variety and information overload. The causal relationship analysis revealed that perceived variety positively influenced hedonic and utilitarian shopping value. However, information overload had a negative effect on hedonic and utilitarian shopping value. Both the hedonic and utilitarian shopping value positively influenced attitudes towards the sites. This study demonstrates that classification method influences customer perception and attitude. It offers interesting insights on a product classification method as a strategic tool for online shopping.

• 한국정밀공학회지
• /
• 제19권11호
• /
• pp.103-111
• /
• 2002

In this study, the natural flaws in welding parts are classified using the signal pattern classification method. The storage digital oscilloscope including FFT function and enveloped waveform generator is used and the signal pattern recognition procedure is made up the digital signal processing, feature extraction, feature selection and classifier design. It is composed with and discussed using the distance classifier that is based on euclidean distance the empirical Bayesian classifier. feature extraction is performed using the class-mean scatter criteria. The signal pattern classification method is applied to the signal pattern recognition of natural flaws.

• 한국농공학회논문집
• /
• 제58권6호
• /
• pp.79-92
• /
• 2016

The objective of this study is to classify the types of rural areas (138 $si{\cdot}gun$) considering settlement environment by Decision Tree Method (CHAID). The CHAID method was used for decision tree algorithm and the seven dependant variables and 5 explanatory variables were selected, respectively. By decision tree method, rural areas were finally classified into six groups through three separate processes. City area, lower area in aging rate and higher area in farmland area ratio was analyzed to be relatively rich rather than other area in the case of settlement environment index. In the future, this study will be able to utilize as a reference to the planning of rural development projects.

• 한국컴퓨터정보학회:학술대회논문집
• /
• 한국컴퓨터정보학회 2019년도 제60차 하계학술대회논문집 27권2호
• /
• pp.267-270
• /
• 2019

Through this research, the rating data of shops were analyzed. The model was designed for discrete multiple classification as to the corresponding data, and the following experiments were initiated to observe the learned machine. By comparing each benchmarks in the experiments, which contains different setting variables for the machine model, the hit ratio was measured which indicates how much it is matched with the expected label. By analyzing those results from each benchmarks, the model was redesigned one time during the research and the effects of each setting variables on this machine were clarified. Furthermore, the research result left the future works, which are related with how the learning could be improved and what should be designed in the further research.