• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.2
• /
• pp.102-112
• /
• 2012

In this paper, we propose a texture classification using wavelet-domain BDIP (block difference of inverse probabilities) and BVLC (block variance of local correlation coefficients) features with WPCA (whitened principal component analysis) classifier. In the proposed method, the wavelet transform is first applied to a query image. The BDIP and BVLC operators are next applied to the wavelet subbands. Global moments for each subband of BDIP and BVLC are then computed and fused into a feature vector. In classification, the WPCA classifier, which is usually adopted in the face identification, searches the training feature vector most similar to the query feature vector. Experimental results show that the proposed method yields excellent texture classification with low feature dimension for test texture image DBs.

• The KIPS Transactions:PartD
• /
• v.19D no.4
• /
• pp.263-270
• /
• 2012

With the current development of high speed Internet and massive database technology, the amount of web documents increases rapidly, and thus, classifying those documents automatically is getting important. In this study, we propose an effective method to extract document features based on Hangeul morpheme and keyword analyses, and to classify non-structured documents automatically by predicting subjects of those documents. To extract document features, first, we select terms using a morpheme analyzer, form the keyword set based on term frequency and subject-discriminating power, and perform the scoring for each keyword using the discriminating power. Then, we generate the classification model by utilizing the commercial software that implements the decision tree, neural network, and SVM(support vector machine). Experimental results show that the proposed feature extraction method has achieved considerable performance, i.e., average precision 0.90 and recall 0.84 in case of the decision tree, in classifying the web documents by subjects.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.6 no.3
• /
• pp.14-24
• /
• 1996

A fuzzy classifier which needs various analyses of features using genetic algorithms is proposed. The fuzzy classifier has a simple structure, which contains a classification part based on fuzzy logic theory and a rule generation ation padptu sing genetic algorithms. The rule generation part determines optimal fuzzy membership functions and inclusior~ or exclusion of each feature in fuzzy classification rules. We analyzed recognition rate of a specific object, then added finer features repetitively, if necessary, to the object which has large misclassification rate. And we introduce repetitive analyses method for the minimum size of string and population, and for the improvement of recognition rates. This classifier is applied to three examples of the classification of iris data, the discrimination of thyroid gland cancer cells and the recognition of confusing handwritten and printed numerals. In the recognition of confusing handwritten and printed numerals, each sample numeral is classified into one of the groups which are divided according to the sample structure. The fuzzy classifier proposed in this paper has recognition rates of 98. 67% for iris data, 98.25% for thyroid gland cancer cells and 96.3% for confusing handwritten and printed numeral!;.

• Journal of the Korea Convergence Society
• /
• v.12 no.8
• /
• pp.39-45
• /
• 2021

This paper discusses how to handle incomplete data including missing values. Optimally processing the missing value means obtaining an estimate that is the closest to the original value from the information contained in the training data, and replacing the missing value with this value. The way to achieve this is to use a decision tree that is completed in the process of classifying information by the classifier. In other words, this decision tree is obtained in the process of learning by inputting only complete information that does not include loss values among all training data into the C4.5 classifier. The nodes of this decision tree have classification variable information, and the higher node closer to the root contains more information, and the leaf node forms a classification region through a path from the root. In addition, the average of classified data events is recorded in each region. Events including the missing value are input to this decision tree, and the region closest to the event is searched through a traversal process according to the information of each node. The average value recorded in this area is regarded as an estimate of the missing value, and the compensation process is completed.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.8
• /
• pp.365-370
• /
• 2023

This paper aims to compare the performance of speech data classification into two groups, adult and elderly, based on the acoustic and linguistic characteristics that change due to aging, such as changes in respiratory patterns, phonation, pitch, frequency, and language expression ability. For acoustic features we used attributes related to the frequency, amplitude, and spectrum of speech voices. As for linguistic features, we extracted hidden state vector representations containing contextual information from the transcription of speech utterances using KoBERT, a Korean pre-trained language model that has shown excellent performance in natural language processing tasks. The classification performance of each model trained based on acoustic and linguistic features was evaluated, and the F1 scores of each model for the two classes, adult and elderly, were examined after address the class imbalance problem by down-sampling. The experimental results showed that using linguistic features provided better performance for classifying adult and elderly than using acoustic features, and even when the class proportions were equal, the classification performance for adult was higher than that for elderly.

• Proceedings of the KSRS Conference
• /
• 2007.03a
• /
• pp.49-54
• /
• 2007

고분해능복사계(AVHRR) 자료로부터 산출한 아시아지역 지면피복 분류자료들 (United States Geological Survey: USGS, International Geosphere Biosphere Programme: IGBP, University of Maryland: UMd)의 분류특성을 분석하였으며 이를 근거로 하여 이 지역에 대한 지면피복의 분류를 시도하였다. 서로 다른 지면피복 분류 자료들의 비교를 위하여 지도 투영법을 일치시켰으며 지면피복 정의가 유사한 유형들만 비교하였다. 세 지면피복 자료에서 분류가 모두 일치하는 비율은 33.57%이고 3 자료 중 두 자료에서 분류가 일치하는 비율은 49.69%로 나타났다. 전체적으로 나대지(사막)， 도시 및 혼합림과 같이 식생의 생물리적 특성이 뚜렷한 유형들에서는 분류의 일치율이 높게 나타났다. 반면에 농지， 낙엽활엽수림， 및 낙엽침엽수렴과 같이 식생의 생물리적 특성이 유사한 유형에서는 일치율이 낮게 나타났다. 분류에 사용된 기본 입력자료수， 지면피복 유형수，분류기법 및 입력 자료의 전처리 수준 등이 지면피복 분류 결과에 차이를 유발한 것으로 판단된다. 지면피복 자료들의 비교결과와 각 유형별 식생지수의 평균 계절변동 특성을 이용하여 이 지역에 대한 지면피복 분류자료를 보완하였다.

• Korean Journal of Plant Taxonomy
• /
• v.37 no.3
• /
• pp.255-273
• /
• 2007

The morphological characters for 4 taxa of Korean Symplocaceae were examined. Based on morphological examination, numerical analysis was performed to clarify the taxonomic relationships among the taxa. It was found that epidermal characters of leaves, cellular deposition of petals, cellular boundary of petioles, growth of stigmas and pollen grains and their surfaces were useful diagonostic characters. From the results of PCA analysis, four taxa were grouped as species clusters including each populations. Four species of Symplocos were clustered as species groups with clear delimitation. Morphological, numerical, and palynological analyses supported the previous morphological studies on this family.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10b
• /
• pp.57-59
• /
• 1999

본 논문에서는 최대 엔트로피 원리에 기반한 문서 분류기의 학습을 제안한다. 최대 엔트로피 기법은 자연언어 처리에서 언어 모델링(Language Modeling), 품사 태깅 (Part-of-Speech Tagging) 등에 널리 사용되는 방법중의 하나이다. 최대 엔트로피 모델의 효율성을 위해서는 자질 선정이 중요한데, 본 논문에서는 자질 집합의 선택을 위한 기준으로 chi-square test, log-likelihood ratio, information gain, mutual information 등의 방법을 이용하여 실험하고, 전체 후보 자질에 대한 실험 결과와 비교해 보았다. 데이터 집합으로는 Reuters-21578을 사용하였으며, 각 클래스에 대한 이진 분류 실험을 수행하였다.

• Proceedings of the IEEK Conference
• /
• 2003.07d
• /
• pp.1499-1502
• /
• 2003

본 논문에서는 손 자세 인식을 위해 손 영상을 소그룹으로 나누고 최종적으로 소그룹 내에서 개별 모델로 분류하는 다단계 접근 방식을 취한다. 이 방식은 처음부터 모든 특성치들을 다 구하여 기존에 가지고 있는 모델 모두와 비교하는 대신, 먼저 소그룹으로 분류 후에 해당 소그룹 내의 모델만을 대상으로 비교 연산을 수행한다. 따라서 계산 량을 크게 줄일 수 있을 뿐 아니라, 확장이 용이하며, 각 소그룹 별로 특성화된 처리를 할 수 있으므로 효율적인 인식기의 구현이 가능하다.

• Proceedings of the Korean Statistical Society Conference
• /
• 2000.11a
• /
• pp.165-171
• /
• 2000

일원분류 모형에서 표준 F-검정을 하기 위해서는 오차항에 대한 등분산성을 가정한다. 그러나 실제로 이러한 가정은 지켜지기 힘들며, 이에 더불어 관찰치가 각 집단별로 일정하지 않고 불균형한 경우에는 F-검정의 유의수준이 지정된 값을 만족시키지 못하며, 따라서 검정력에 관한 분석은 의미가 없게 된다. 본 연구에서는 등분산성이 지켜지지 않고, 자료가 불균형한 경우, 현실적인 상황에서 일반적으로 사용되는 F-검정의 유의수준 유지라는 문제가 어 떤 변화를 겪게 되는지를 확인하고, 더 나아가 유의수준을 유지하기 위해서는 어떤 식의 조정이 가능한지를 살펴보았다.