• 대한원격탐사학회지
• /
• 제37권4호
• /
• pp.719-731
• /
• 2021

This study proposes a two-stage hybrid classification model for crop classification using multi-temporal remote sensing images; the model combines feature embedding by using an autoencoder (AE) with a convolutional neural network (CNN) classifier to fully utilize features including informative temporal and spatial signatures. Long short-term memory (LSTM)-based AE (LAE) is fine-tuned using class label information to extract latent features that contain less noise and useful temporal signatures. The CNN classifier is then applied to effectively account for the spatial characteristics of the extracted latent features. A crop classification experiment with multi-temporal unmanned aerial vehicle images is conducted to illustrate the potential application of the proposed hybrid model. The classification performance of the proposed model is compared with various combinations of conventional deep learning models (CNN, LSTM, and convolutional LSTM) and different inputs (original multi-temporal images and features from stacked AE). From the crop classification experiment, the best classification accuracy was achieved by the proposed model that utilized the latent features by fine-tuned LAE as input for the CNN classifier. The latent features that contain useful temporal signatures and are less noisy could increase the class separability between crops with similar spectral signatures, thereby leading to superior classification accuracy. The experimental results demonstrate the importance of effective feature extraction and the potential of the proposed classification model for crop classification using multi-temporal remote sensing images.

• 한국멀티미디어학회논문지
• /
• 제14권9호
• /
• pp.1190-1201
• /
• 2011

본 연구는 간호사 및 간호학생을 위한 간호진단, 간호중재, 간호결과 분류체계의 간호과정 프로그램을 모바일 응용 기반으로 개발하였다. 연구재료는 표준화된 분류체계인 북미간호진단협의회의 간호진단 분류체계와 아이오와 대학을 중심으로 개발된 간호중재 분류체계, 간호결과 분류체계를 사용하였다. 기존 연구 방법은 간호과정의 일부분만을 선택하여 개발하므로 환경에 제한적인 프로그램으로 임상에 일반화시켜 환자들에게 적용하기 어려웠다. 그러나 본 연구는 진단-결과-중재의 전체를 연계시킨 프레임워크를 개발하므로 어떠한 임상환경에서도 적용이 가능한 가이드라인으로 개발하였다. 개발된 프로그램은 한글판으로 3월부터 앱 스토어에 등록되었으며 간호교육 도구로 적극적으로 활용되기를 기대한다.

• 한국멀티미디어학회논문지
• /
• 제21권5호
• /
• pp.617-625
• /
• 2018

Phishing website has become a crucial concern in cyber security applications. It is performed by fraudulently deceiving users with the aim of obtaining their sensitive information such as bank account information, credit card, username, and password. The threat has led to huge losses to online retailers, e-business platform, financial institutions, and to name but a few. One way to build anti-phishing detection mechanism is to construct classification algorithm based on machine learning techniques. The objective of this paper is to compare different classifier ensemble approaches, i.e. random forest, rotation forest, gradient boosted machine, and extreme gradient boosting against single classifiers, i.e. decision tree, classification and regression tree, and credal decision tree in the case of website phishing. Area under ROC curve (AUC) is employed as a performance metric, whilst statistical tests are used as baseline indicator of significance evaluation among classifiers. The paper contributes the existing literature on making a benchmark of classifier ensembles for web phishing detection.

• 한국지능시스템학회논문지
• /
• 제11권7호
• /
• pp.633-640
• /
• 2001

In this paper we examine the machine learning issues raised by the domain of the Intrusion Detection Systems(IDS), which have difficulty successfully classifying intruders. There systems also require a significant amount of computational overhead making it difficult to create robust real-time IDS. Machine learning techniques can reduce the human effort required to build these systems and can improve their performance. Genetic algorithms are used to improve the performance of search problems, while data mining has been used for data analysis. Data Mining is the exploration and analysis of large quantities of data to discover meaningful patterns and rules. Among the tasks for data mining, we concentrate the classification task. Since classification is the basic element of human way of thinking, it is a well-studied problem in a wide variety of application. In this paper, we propose a classifier system based on genetic algorithm, and the proposed system is evaluated by applying it to IDS problem related to classification task in data mining. We report our experiments in using these method on KDD audit data.

• Journal of Multimedia Information System
• /
• 제5권2호
• /
• pp.99-104
• /
• 2018

Phishing website has become a crucial concern in cyber security applications. It is performed by fraudulently deceiving users with the aim of obtaining their sensitive information such as bank account information, credit card, username, and password. The threat has led to huge losses to online retailers, e-business platform, financial institutions, and to name but a few. One way to build anti-phishing detection mechanism is to construct classification algorithm based on machine learning techniques. The objective of this paper is to compare different classifier ensemble approaches, i.e. random forest, rotation forest, gradient boosted machine, and extreme gradient boosting against single classifiers, i.e. decision tree, classification and regression tree, and credal decision tree in the case of website phishing. Area under ROC curve (AUC) is employed as a performance metric, whilst statistical tests are used as baseline indicator of significance evaluation among classifiers. The paper contributes the existing literature on making a benchmark of classifier ensembles for web phishing detection.

• 통상정보연구
• /
• 제6권3호
• /
• pp.41-61
• /
• 2004

Today, most companies are needed to become interested on e-Biz and information management system. Especially, Data classification format system was very important for application to effective and efficiency management decision support. They should include main entry which consists of department, employee's name, title, publication date. Now, each company is using eleven different methods on data classification format system. In this paper finding result was as follows, in other words, general management document case using the nine date classification methods and special report management document ca se using the twodata classification methods. The aim of this study is to investigate problems that the present data classification format system has and some concerns that should be taken into account in case of the modification of the data classification system and change into a new one. This study is based on the survey in that the company managergave to 35 companies throughout the nation. As a result, the survey indicates that the crucial concerns of the participating managers are ineffective management information source and the duplication of data classification systems. This paper is the transcendental study the introduction of data classification format systems to business companies in Korea. This paper provided the fundamental data for the effective business process reengineering in business activity for management information.

• 전기학회논문지
• /
• 제59권2호
• /
• pp.277-284
• /
• 2010

As technology is developed, the quantity of new vocabularies is increasing more rapidly. So many vocabularies of technology have various meanings for each part and are used diversely according to circumstances. Therefore, the necessity of reasonable methods of standardization and purification is increasing and it is necessary to establish a classification system of terminology for the first phase of the standardization. Firstly, based on classification systems of power and IT standard dictionaries, scientific and technological standard, SPARK, power IT fields of IEC and organization units of corporations, we propose a new classification system for the standardization of power IT terminologies. The classification system consists of a hierarchical structure with general classification, application fields and specific technologies while keeping the conventional matrix-type classification system. And interpretation methods of power IT terminologies, which are classified according to the new classification system for the standardization of power IT terminologies, is proposed. The interpretation works of the power IT terminologies confirm that the classification system is systematic and the interpretation process is efficient.

• 한국통신학회논문지
• /
• 제36권2B호
• /
• pp.131-140
• /
• 2011

최근 인터넷 사용자의 증가와 고속 네트워크 망을 통한 네트워크 트래픽의 급증으로 효율적인 네트워크 트래픽 관리의 필요성이 더욱 커졌다. 효율적인 트래픽 관리를 위해서는 응용 프로그램 별 트래픽 분류의 연구가 선행되어야 하며 이미 많은 기존 논문에서 응용레벨 트래픽 분류에 대한 다양한 알고리즘을 제시하고 있다. 하지만 P2P기반의 Skype응용에 대해서는 분석율이 떨어져 이에 대한 연구가 더 필요한 실정이다. 본 논문에서는 payload 시그니쳐 기반 분석, 기계학습 기반 분석 등 기존의 방법론에 의존하지 않고 Skype응용의 트래픽 특성을 분석해 사용자들의 {IP, port} 리스트를 추출하고 이를 이용해 네트워크 내에 발생하는 Skype응용 프로그램의 트래픽을 정확하게 탐지하는 실시간 탐지 알고리즘을 제안한다 제안된 방법론은 학내 네트워크에 적용하여 그 타당성을 검증하였다.

• 경영과학
• /
• 제11권1호
• /
• pp.129-144
• /
• 1994

B-ISDN(Broadband Integrated Services Digital Network)which is defined as a service or system requiring transmission channels capable of supporting rates above 1.5 Mbps has emerged as a new future telecommunication infrastructure. B-ISDN can integrate a wide range of services and the success of B-ISDN is crucially dependent on the development of user-needing application services. The purpose of this study is the conceptual development of B-ISDN application services. We survey on the kinds of B-ISDN service, classify application areas by user groups, and develop B-ISDN application services. Finally we categorize B-ISDN application services by their application areas and necessary services.

• 한국멀티미디어학회논문지
• /
• 제16권7호
• /
• pp.795-801
• /
• 2013

서포트 벡터 머신은 뛰어난 일반화 성능에 힘입어 다양한 분야에서 의사 결정 나무나 인공 신경망에 비해 더 좋은 분류 성능을 보이고 있기 때문에 최근 널리 사용되고 있다. 서포트 벡터 머신은 기본적으로 이진 분류 문제를 위하여 설계되었기 때문에 서포트 벡터 머신을 다중 클래스 문제에 적용하기 위한 방법으로 다중 이진 분류기의 출력 결과를 이용하는 출력 코딩 방법이 주로 사용되고 있다. 그러나 출력 코딩 기반 서포트 벡터 머신에 사용된 기존 특징 선택 기법은 각 분류기의 정확도 향상을 위한 특징이 아니라 전체 분류 정확도 향상을 위한 특징을 선택하고 있다. 본 논문에서는 출력 코딩 기반 서포트 벡터 머신의 각 이진 분류기의 분류 정확도를 최대화하는 특징을 각각 선택하여 사용함으로써, 전체 분류 정확도를 향상시키는 특징 선택 기법을 제안한다. 실험 결과는 제안 기법이 기존 특징 선택 기법에 비하여 통계적으로 유의미한 분류 정확도 향상이 있었음을 보여주었다.