• The KIPS Transactions:PartB
• /
• v.16B no.3
• /
• pp.233-238
• /
• 2009

This paper presents an approach to the classifications of small size document using the instance-based feature filtering Relief-F algorithm. In the document classifications, we have not always good classification performances of small size document included a few features. Because total number of feature in the document set is large, but feature count of each document is very small relatively, so the similarities between documents are very low when we use general assessment of similarity and classifiers. Specially, in the cases of the classification of web document in the directory service and the classification of the sectors that cannot connect with the original file after recovery hard-disk, we have not good classification performances. Thus, we propose the Extended Relief-F(ERelief-F) algorithm using instance-based feature filtering algorithm Relief-F to solve problems of Relief-F as preprocess of classification. For the performance comparison, we tested information gain, odds ratio and Relief-F for feature filtering and getting those feature values, and used kNN and SVM classifiers. In the experimental results, the Extended Relief-F(ERelief-F) algorithm, compared with the others, performed best for all of the datasets and reduced many irrelevant features from document sets.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.6
• /
• pp.654-662
• /
• 2010

In this paper, we suggest an advanced algorithm, to recognize pedestrian/non-pedestrian using second-stage cascade method, which applies Adaboost algorithm to make a strong classification from weak classifications. First, we extract two feature vectors: (i) Histogram of Oriented Gradient (HOG) which includes gradient information and differential magnitude; (ii) Curvature-HOG which is based on four different curvature features per pixel. And then, a strong classification needs to be obtained from weak classifications for composite recognition method using both HOG and curvature-HOG. In the proposed method, we use one feature vector and one strong classification for the first stage of recognition. For the recognition-failed image, the other feature and strong classification will be used for the second stage of recognition. Based on our experiment, the proposed algorithm shows higher recognition rate compared to the traditional method.

• Journal of Korean Society of Transportation
• /
• v.27 no.4
• /
• pp.55-61
• /
• 2009

In order to observe the impact of a type of vehicles for traffic flows and pavement, vehicle classifications is conducted. Korean Ministry of Land, Transport and Maritime Affairs provides 12-type vehicle classifications on National expressways, National highways, and Provincial roads. Current AVC (Automatic Vehicle Classification) devices decide vehicle types comparing measurements of vehicle lengths, wheelbases, overhangs etc. to a reference table including those of all types of models. This study developed an algorithm for macroscopic vehicle classification which is less sensitive to tuning sensors and updating the reference table. For those characteristics, trend lines in vehicle lengths and wheelbases are employed. To assess the algorithm developed, vehicle lengths and wheelbases were collected from an AVC device. In this experiment, this algorithm showed the accuracy of 88.2 % compared to true values obtained from video replaying. Our efforts in this study are expected to contribute to developing devices for macroscopic vehicle classification.

• Journal of Chest Surgery
• /
• v.24 no.1
• /
• pp.72-82
• /
• 1991

From May 1978 to Sep. 1990, 106 patients who had been diagnosed as primary lung cancer and operated on at the Department of Thoracic & Cardiovascular Surgery, Han Yang University, were clinically evaluated. 1. The peak incidence of age was 5th decade of life[37.7%] and 6th decade[29.2%]. Male to female ratio was 3.8: l. 2. Most of symptoms were respiratory, which were cough, chest pain, hemoptysis, and asymptomatic cases were 2.9%. 3. Histopathologic classifications were squamous cell carcinoma[53.7%], adenocarcinoma [23.8%], bronchioloalveolar cell carcinoma[6.6%], undifferentiated large cell carcinoma[6.6%], small cell carcinoma[3.8%], adenosquamous carcinoma[3.8%] and others[1.8%]. 4. Methods of operation were pneumonectomy 49.1%[52cases], lobectomy 21%[22cases] bilobectomy[6cases], lobectomy with wedge resection[3cases], exploration 21.9%[23cases], and resectability was 78.3%. 5. Staging classifications were Stage I [22.6%], Stage II [11.3%], Stage IIIa[42.6%], Stage IIIb[21.7%] and Stage lV[1.6%]. Resectability by Stage; Stage I was 100%, II 100%, IIIa 84.4% and IIIb 30.4%. 6. Causes of most of inoperable cases were invasion of mediastinal structures and diffuse chest wall, and others were contralateral lymph node invasion and malignant pleural effusion. 7. Operative mortality was 6.7% which caused by arrhythmia, sepsis, pulmonary edema, and radiation pneumonitis. 8. On the long term follow up of the resectable cases, overall 1 year survival rate was 58.5 %, 2 year 39%, and 5 year 19.5%. Five year survival rate was 40% in Stage I, 25% in Stage II and 11.7% in Stage Illa. As for the method of operation, the higher 5 year survival rate was observed in lobectomies[33.3%] than in pneumonectomies[10.3%].

• Journal of Digital Convergence
• /
• v.12 no.5
• /
• pp.115-123
• /
• 2014

Patent information analysis has been carried out for technological capability analysis of competitors relating to next generation mobile communication. Various analysis methods, such as applicant analysis, technology classification analysis, indicator analysis and the like have been utilized as a method of analyzing patent information. As a first step for the technological capability analysis of competitors, applicants having high patent activity(PA) were selected, and as a second step therefor, technology classifications showing high technological independence (TI) were selected. Furthermore, portfolios for technology classifications showing high technological independence in the patents of main applicants having high patent activity by matching results the first and second steps together were prepared. Through such a process, portfolios for important technologies which have been concentrically researched by competitors could be analyzed. Accordingly, the present analysis results will help to carry out strategic R&D management, such as the establishment of company R&D plans and patent strategies.

• Fire Science and Engineering
• /
• v.23 no.4
• /
• pp.112-119
• /
• 2009

Building use classification is classified the building types according to the similar structure, the goal of the use, and its shape. The important reason of building use classification is that building restrictions are applied according to building use classifications. But there are a lot of confusion because the building use classification is regulated not only by the Building Act but also by the several individual Act. Particularly in the large-scaled, high-raised and complex building construction, it is very important to remain the consistency of the building use classification between the Building Act and the Fire-Fighting System Act for safety of occupants. So the purpose of this study is to suggests the improve direction of the building use classifications through the comparative analysis of the Building Act and the Fire-Fighting System Act.

• Journal of the Korean Society for information Management
• /
• v.27 no.4
• /
• pp.49-69
• /
• 2010

Library classification system is based upon academic disciplines, However, it is difficult to classify for Internet resources due to its lack of up-to-datedness and practicality. Especially, headings of Korean Decimal Classification need to reflect practical aspects and it should be also developed for classification of web based resources. The purposes of this study are to analyze the structures of directory classifications in Internet resources and to suggest additional headings of KDC as a practical library classification as well as a classification system for internet resources. Directory classification systems of Naver, Yahoo, Kyobo Internet book store, Amazon were selected and their food and culture subjects were analyzed for this study. The headings of KDC were compared to them and new possible headings were suggested with reference of NDC and DDC in the area of food and culture. This study provided a way of developing KDC for a classification system for Internet resources as well as library materials.

• Journal of Preventive Medicine and Public Health
• /
• v.39 no.4
• /
• pp.293-301
• /
• 2006

The preventive medicine learning objectives, first developed in 1977 and subsequently supplemented, underwent necessary revision of the contents for the fourth time to create the fifth revision. However, the required educational contents of health promotion and disease prevention have been changed by the new trends of medical education such as PBL and integrated curriculum, the rapid change of the health and medical environment and the globalization of medicine. The Korean Society of Preventive Medicine formed a task force, led by the Undergraduate Education Committee in 2003, which surveyed all the medical colleges to describe the state of preventive medicine education in Korea, analyzed the changing education demand according to the change of health environment and quantitatively measured the validity and usefulness of each learning objective in the previous curriculum. Based on these data, some temporary objectives were formed and promulgated to all the medical schools. After multiple revisions, an almost completely new series of learning objectives for preventive medicine was created. The objectives comprised 4 classifications and 1 supplement: 1) health and disease, 2) epidemiology and its application, 3) environment and health, 4) patient-doctor-society, and supplementary clinical occupational health. The total number of learning objectives, contained within 13 sub-classifications, was 221 (including 35 of supplementary clinical occupational health). Future studies of the learning process and ongoing development of teaching materials according to the new learning objectives should be undertaken with persistence in order to ensure the progress of preventive medicine education.

• The Korean Journal of Applied Statistics
• /
• v.32 no.2
• /
• pp.265-276
• /
• 2019

The document-term frequency matrix is a general data of objects in text mining. In this study, we introduce a traditional term weighting scheme TF-IDF (term frequency-inverse document frequency) which is applied in the document-term frequency matrix and used for text classifications. In addition, we introduce and compare TF-IDF-ICSDF and TF-IGM schemes which are well known recently. This study also provides a method to extract keyword enhancing the quality of text classifications. Based on the keywords extracted, we applied support vector machine for the text classification. In this study, to compare the performance term weighting schemes, we used some performance metrics such as precision, recall, and F1-score. Therefore, we know that TF-IGM scheme provided high performance metrics and was optimal for text classification.

• Korea Trade Review
• /
• v.41 no.5
• /
• pp.163-185
• /
• 2016

The impact of global trade on developing countries is a critical subject. Especially in this country, under difficult conditions like political issues with North Korea, achieved a trade surplus of 90.2 billion dollars last year. While the service trade recorded a deficit of 15.7 billion dollars in the same period. According to WTO, services are either the result of a production activity that changes the conditions of the consuming units (transformation services), or facilitates the exchange of products or financial assets (margin services). To unify and compare with the trade and service statistics of countries, they assorted 12 classifications of services like Manufacturing service on physical inputs owned by others, Maintenance and repair service n.i.e., Transport, Travel, Construction, Insurance and pension service, Financial service, Charges for the use of intellectual property n.i.e., Telecommunications, Computer and information services, Other business services, Personal, cultural and recreational services, and Government goods and services n.i.e. Thus, this study is to estimate the international competitiveness of service trade between 2006 and 2015 in Republic of Korea, according to EBOPS 6th. As a result, total service volume of export and import has increased in the last 10 years. Its volume, recently, accounts for 22% of total goods trade. Also the Korean Government can make a supporting policy and decide a supporting business in terms of service trade. Finally, we can find a mutual connectivity within 12 service businesses and between Service and Goods.