• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.5
• /
• pp.341-346
• /
• 2013

In this paper, we propose a rule extraction method using a modified Fuzzy Min-Max (FMM) neural network. The suggested method supplements the hyperbox definition with a frequency factor of feature values in the learning data set. We have defined a relevance factor between features and pattern classes. The proposed model can solve the ambiguity problem without using the overlapping test process and the contraction process. The hyperbox membership function based on the fuzzy partitions is defined for each dimension of a pattern class. The weight values are trained by the feature range and the frequency of feature values. The excitatory features and the inhibitory features can be classified by the proposed method and they can be used for the rule generation process. From the experiments of sign language recognition, the proposed method is evaluated empirically.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.14 no.1
• /
• pp.16-22
• /
• 2010

Electrocardiogram(ECG) analysis and arrhythmia recognition are critical for diagnosis and treatment of ill patients. Cardiac arrhythmia is a condition in which heart beat may be irregular and presents a serious threat to the patient recovering from ventricular tachycardia (VT) and ventricular fibrillation (VF). Other arrhythmias like atrial premature contraction (APC), Premature ventricular contraction (PVC) and superventricular tachycardia (SVT) are important in diagnosing the heart diseases. This paper presented new method to classify various arrhythmias contrary to other techniques which are limited to only two or three arrhythmias. ECG is decomposed into Intrinsic Mode Functions (IMFs) by Empirical Mode Decomposition (EMD). Burg algorithm was performed on IMFs to obtain AR coefficients which can reduce the dimension of feature vector and utilized as Multi-class SVM inputs which is basically extended from binary SVM. We chose optimal parameters for SVM classifier, applied to arrhythmias classification and achieved the accuracies of detecting NSR, APC, PVC, SVT, VT and VP were 96.8% to 99.5%. The results showed that EMD was useful for the preprocessing and feature extraction and multi-class SVM for classification of cardiac arrhythmias, with high usefulness.

• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.48 no.5
• /
• pp.31-36
• /
• 2011

In this paper, Mel-Frequency Cepstral Coefficient (MFCC), Decorrelated Filter Bank (DFB), Octave-based Spectral Contrast (OSC), Zero-Crossing Rate (ZCR), and Spectral Contract/Roll-Off are combined as a set of multiple feature-vectors for the music genre classification system based on the Support Vector Machine (SVM) classifier. In the conventional system, feature vectors for the entire genre classes are normalized for the SVM model training and classification. However, in this paper, selected feature vectors that are compared based on the One-Against-One (OAO) SVM classifier are only used for normalization. Using OSC as a single feature-vector and the multiple feature-vectors, we obtain the genre classification rates of 60.8% and 77.4%, respectively, with the conventional normalization method. Using the proposed normalization method, we obtain the increased classification rates by 8.2% and 3.3% for OSC and the multiple feature-vectors, respectively.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.05a
• /
• pp.44-47
• /
• 2008

본 논문에서는 유비쿼터스 환경에서 인간의 행동패턴을 인식하고 이 결과를 활용하여 사용자의 의도를 추론하는 방법론에 관해 기술한다. 인간행동의 예측에 관한 지식표현으로부터의 추론기능과 예제패턴 기반의 학습기능을 동시에 지원하는 모델을 제시하고 이론의 타당성과 유용성을 고찰한다. 의도 추론 문제에서 지식기반 기법이 갖는 불완전성을 극복하기 위하여 예제기반 학습능력의 필요성을 도출하는 한편, 다양한 변이가 존재하는 응용에서 학습데이터 선정의 어려움을 보완하기 위한 방법론을 제시한다. 세부적으로 인간행동에 관한 특징표현과 행동패턴 클래스를 정의하고 이들간의 관계를 고유한 지식표현 규칙으로 정형화 한다. 또한 제안된 지식표현을 수용하는 추론 메커니즘을 제시하며, 제시한 모델의 부수적 특징으로 학습과정을 통한 지식 정련기능의 유용성을 고찰한다.

• Journal of Broadcast Engineering
• /
• v.13 no.6
• /
• pp.828-837
• /
• 2008

We propose a method to select representative views of single objects and classes of objects for 3D object retrieval and classification. Our method is based on projected 2D shapes, or views, of the 3D objects, where the representative views are selected by applying affinity propagation to cluster uniformly sampled views. Affinity propagation assigns prototypes to each cluster during the clustering process, thereby providing a natural criterion to select views. We recursively apply affinity propagation to the selected views of objects classified as single classes to obtain representative views of classes of objects. By enabling classification as well as retrieval, effective management of large scale databases for retrieval can be enhanced, since we can avoid exhaustive search over all objects by first classifying the object. We demonstrate the effectiveness of the proposed method for both retrieval and classification by experimental results based on the Princeton benchmark database [16].

• Journal of KIISE:Software and Applications
• /
• v.32 no.1
• /
• pp.1-9
• /
• 2005

In this paper we propose a modified fuzzy min-max neural network model for pattern classification and discuss the usefulness of the model. We define a new hypercube membership function which has a weight factor to each of the feature within a hyperbox. The weight factor makes it possible to consider the degree of relevance of each feature to a class during the classification process. Based on the proposed model, a knowledge extraction method is presented. In this method, a list of relevant features for a given class is extracted from the trained network using the hyperbox membership functions and connection weights. Ft)r this purpose we define a Relevance Factor that represents a degree of relevance of a feature to the given class and a similarity measure between fuzzy membership functions of the hyperboxes. Experimental results for the proposed methods and discussions are presented for the evaluation of the effectiveness and feasibility of the proposed methods.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.9
• /
• pp.375-384
• /
• 2021

Deep learning techniques have been proven to have high performance in image processing and are applied in various fields. The most widely used methods for validating a deep learning model include a holdout verification method, a k-fold cross verification method, and a bootstrap method. These legacy methods consider the balance of the ratio between classes in the process of dividing the data set, but do not consider the ratio of various features that exist within the same class. If these features are not considered, verification results may be biased toward some features. Therefore, we propose a deep learning model validation method based on data feature coverage for image classification by improving the legacy methods. The proposed technique proposes a data feature coverage that can be measured numerically how much the training data set for training and validation of the deep learning model and the evaluation data set reflects the features of the entire data set. In this method, the data set can be divided by ensuring coverage to include all features of the entire data set, and the evaluation result of the model can be analyzed in units of feature clusters. As a result, by providing feature cluster information for the evaluation result of the trained model, feature information of data that affects the trained model can be provided.

• The Journal of Bigdata
• /
• v.8 no.2
• /
• pp.177-187
• /
• 2023

Semi-supervised learning (SSL) is an effective approach to training models using a small amount of labeled data and a larger amount of unlabeled data. However, many papers in the field use a fixed threshold when applying pseudo-labels without considering the feature-wise differences among images of different classes. In this paper, we propose a SSL method for synthetic aperture radar (SAR) image classification that applies different thresholds for each class instead of using a single fixed threshold for all classes. We propose a threshold learning module into the model, considering the differences in feature distributions among classes, to dynamically learn thresholds for each class. We compare the application of a SSL SAR image classification method using different thresholds and examined the advantages of employing class-specific thresholds.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.45 no.4
• /
• pp.19-26
• /
• 2008

Gene set enrichment analysis (GSEA) is a computational method to identify statistically significant gene sets showing significant differences between two groups of microarray expression profiles and simultaneously uncover their biological meanings in an elegant way by employing gene annotation databases, such as Cytogenetic Band, KEGG pathways, gene ontology, and etc. For the gone set enrichment analysis, all the genes in a given dataset are first ordered by the signal-to-noise ratio between the groups and then further analyses are proceeded. Despite of its impressive results in several previous studies, however, gene ranking by the signal-to-noise ratio makes it difficult to consider highly up-regulated genes and highly down-regulated genes at the same time as the candidates of significant genes, which possibly reflect certain situations incurred in metabolic and signaling pathways. To deal with this problem, in this article, we investigate the gene set enrichment analysis method with Fisher criterion for gene ranking and also evaluate its effects in Leukemia related pathway analyses.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.245-246
• /
• 2022

최근 IoT 환경에서 기계학습을 이용한 공격 탐지 모델의 연구가 활발히 진행되고 있으며, 탐지 정확도도 점차 향상되고 있다. 하지만, IoT 환경의 특징인 저 사양 하드웨어, 고차원의 특징, 방대한 트래픽 등으로 인해 탐지성능이 저하되는 문제가 있다. 따라서 본 논문에서는 MQTT(Message Queuing Telementry Transport) 프로토콜 기반의 IoT 환경에서 수집된 데이터셋을 대상으로 주성분 분석(Principal Component Analysis)과 LightGBM을 이용하여 데이터셋 차원을 감소시키고, 공격 클래스를 분류하였다. 실험결과 원본 데이터셋 차원을 주성분 3개(약 9%)로 감소시켰음에도 모든 특징(33개)을 사용한 실험결과와 거의 유사한 성능을 보였다. 또한 기존 연구의 특징 선택을 통한 탐지 모델과 비교하였을 때도 분류성능이 더 우수한 것으로 나타났다.