• Journal of KIISE:Software and Applications
• /
• v.34 no.9
• /
• pp.857-868
• /
• 2007

The good classifier ensemble should have a high complementarity among classifiers in order to produce a high recognition rate and its size is small in order to be efficient. This paper proposes a classifier ensemble selection algorithm with coarse-to-fine stages. for the algorithm to be successful, the original classifier pool should be sufficiently diverse. This paper produces a large classifier pool by combining several different classification algorithms and lots of feature subsets. The aim of the coarse selection is to reduce the size of classifier pool with little sacrifice of recognition performance. The fine selection finds near-optimal ensemble using genetic algorithms. A hybrid genetic algorithm with improved searching capability is also proposed. The experimentation uses the worldwide handwritten numeral databases. The results showed that the proposed algorithm is superior to the conventional ones.

• Journal of KIISE
• /
• v.43 no.1
• /
• pp.80-86
• /
• 2016

A speech-act is a behavior intended by users in an utterance. Speech-act classification is important in a dialogue system. The machine learning and rule-based methods have mainly been used for speech-act classification. In this paper, we propose a speech-act classification method based on the combination of support vector machine (SVM) and transformation-based learning (TBL). The user's utterance is first classified by SVM that is preferentially applied to categories with a low utterance rate in training data. Next, when an utterance has negative scores throughout the whole of the categories, the utterance is applied to the correction phase by rules. The results from our method were higher performance over the baseline system long with error-reduction.

• The Journal of the Korea Contents Association
• /
• v.20 no.8
• /
• pp.359-367
• /
• 2020

WSNs as the main technology of IoT often deliver information or authenticate based on location. Thus, verifying location information is essential. This paper aims to present the comprehensive analysis and classification of location verification techniques in WSN. For this, classification criteria are suggested based on the result of feature analysis of existing techniques. In addition, the existing techniques are classified according to the suggested criteria, and each characteristic and development direction are described. The result of this paper is expected to be a useful reference material when designing a new technique.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.4
• /
• pp.219-226
• /
• 2013

Feature selection is the one of important issues in the field of machine learning and pattern recognition. It is the technique to find a subset from the source data and can give the best classification performance. Ie, it is the technique to extract the subset closely related to the purpose of the classification. In this paper, we experimented to select the best feature subset for improving classification accuracy when classify success and failure factors in software reuse. And we compared with existing studies. As a result, we found that a feature subset was selected in this study showed the better classification accuracy.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04a
• /
• pp.439-441
• /
• 2004

이메일 사용이 보편화됨에 따라 점차 수신되는 메일의 량이 증가하고 있다. 이러한 메일 량의 증가는 사용자로 하여금 이메일을 좀더 효율적으로 분류할 수 있는 방법을 필요하게 한다. 그러나 현재의 이메일 분류는 규칙기반, 베이시안, SVM 등을 이용하여 스팸메일을 필터링 하는 이원분류가 주로 연구되고 있다. 이외에도 다원분류에 대한 연구로는 클러스터링을 이용한 방법이 있으나, 이는 단순히 유사도에 의해 메일을 묶는 수준에 그치고 있다. 본 논문에서는 벡터모델의 유사도를 기반으로 한 분류 주제 자동 생성 알고리즘과 동적분류체계 방법을 결합하여 새로운 이메일 자동 다원분류 방법을 제안했다. 본 논문에서 제안한 방법은 이메일을 자동으로 분류하여, 분류된 결과를 색인검색과 디렉토리 검색 방법을 지원하며 대량의 메일도 효율적으로 관리할 수 있다. 또한 메시지를 동적으로 재분류 할 수 있게 함으로써 디렉토리 검색시 재현율을 높였다.

• Journal of Advanced Marine Engineering and Technology
• /
• v.39 no.1
• /
• pp.58-62
• /
• 2015

Most blog sites provide predefined classes based on contents or topics, but few bloggers choose classes for their posts because of its cumbersome manual process. This paper proposes an automatic blog post classification method that variously combines term frequency, document frequency and class frequency from each classes to find appropriate weighting scheme. In experiment, combination of term frequency, category term frequency and inversed (excepted category's) document frequency shows 77.02% classification precisions.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.3
• /
• pp.614-621
• /
• 2006

We have implemented a multiclassifier learning approach in a GA-based inductive learning environment that learns classification rules that are similar to rules used in PROSPECTOR. In the multiclassifier learning approach, a classification system is constructed with several classifiers that are obtained by running a GA-based learning system several times to improve the overall performance of a classification system. To implement the multiclassifier learning approach, we need a decision-making scheme that can draw a decision using multiple classifiers. In this paper, we introduce two decision-making schemes: one is based on combining posterior odds given by classifiers to each class and the other one is a voting scheme based on ranking assigned to each class by classifiers. We also present empirical results that evaluate the effect of the multiclassifier learning approach on the GA-based inductive teaming environment.

• Proceedings of the Korean Statistical Society Conference
• /
• 2001.11a
• /
• pp.105-108
• /
• 2001

본 논문에서는 결측치가 있는 불완비 데이터에서 분류나루를 구축하는 방법을 고찰하였다. 기존의 결측치 처리 방법인 대리 분리 방법의 대안으로 대체 방법으로 결측치를 처리한 후 분류나무를 구축하는 방법을 제안하였다.

• Korean Journal of Plant Taxonomy
• /
• v.39 no.3
• /
• pp.143-169
• /
• 2009

To confirm the taxonomic identity of taxa described by $L{\acute{e}}veill{\acute{e}}$, H. H. A, a French plant taxonomist, in the serial papers of Decades plantarum novarum based on the collections of Fathers Faurie and Taquet from the Korean peninsula, we examined the numerous references that contained taxonomic opinions about $L{\acute{e}}veill{\acute{e}}^{\prime}s$ taxa. Among them, 146 taxa were confirmed as conspecific with other existing taxa. Of them, 79 taxa, including Ajuga devestita, were listed as a synonym of only one species. Sixty-seven taxa, including Bidens robertianifolia, were listed as synonyms of two or more species according to taxonomists. Eight taxa, including Aconitum coreanum, were considered illegitimate names because of later homonym, or other problems. Five taxa, including Rhododendron hallaisanense, were treated either as distinct or conspecific taxa depending on taxonomists.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04d
• /
• pp.521-523
• /
• 2003

본 논문은 웹 탐색 중 디렉토리 검색엔진의 분류검색에 대한 문제점을 해결하기 위해서 분류와 검색어간의 관계를 퍼지논리를 이용하여 계산하고 분류간의 함의관계를 유도함으로써 동적인 분류체계를 구성하는 새로운 방법을 제시한다. 이 방법의 장점은 분류간의 함의관계를 유사한 하위분류로서 간주함으로써 분류검색 결과의 재현율을 높일 수 있다는 것이다.