• 한국컴퓨터정보학회논문지
• /
• 제18권11호
• /
• pp.31-38
• /
• 2013

특징 정규화는 인식기를 적용하기 이전의 전처리 단계로 특징의 스케일에 따른 오류를 줄이기 위해 널리 사용되고 있다. 하지만 기존 정규화 방법은 특징의 분포를 가정하는 경우가 많으며, 클래스 라벨을 고려하지 않으므로 정규화 결과가 인식률에서 최적임을 보장하지 못하는 문제점이 있다. 이 논문에서는 특징의 분포를 가정하지 않는 랭크 정규화 방법과 클래스 라벨을 사용하는 교사 학습법을 결합한 교사 랭크 정규화 방법을 제안하였다. 제안하는 방법은 데이터의 분포를 바탕으로 특징의 분포를 자동으로 추정하므로 특징의 분포를 가정하지 않으며, 데이터 포인트의 최근접 이웃이 가지는 클래스 라벨을 바탕으로 정규화를 시행하므로 오류의 발생을 최소화할 수 있다. 특히 SVM의 경우 서로 다른 클래스에 속하는 데이터 포인트들이 혼재되어 나타나는 영역에 경계선을 설정하므로 이 영역의 밀도를 줄임으로써 경계선 설정을 보다 용이하게 하고 결과적으로 일반화 오류를 감소시킬 수 있다. 이러한 사실들은 실험 결과를 통해 확인할 수 있다.

• 융합보안논문지
• /
• 제23권1호
• /
• pp.117-123
• /
• 2023

점점 더 고도화되고 있는 랜섬웨어 공격을 기계학습 기반 모델로 탐지하기 위해서는, 분류 모델이 고차원의 특성을 가지는 학습데이터를 훈련해야 한다. 그리고 이 경우 '차원의 저주' 현상이 발생하기 쉽다. 따라서 차원의 저주 현상을 회피하면서 학습모델의 정확성을 높이고 실행 속도를 향상하기 위해 특성의 차원 축소가 반드시 선행되어야 한다. 본 논문에서는 특성의 차원이 극단적으로 다른 2종의 데이터세트를 대상으로 3종의 기계학습 모델과 2종의 특성 추출기법을 적용하여 랜섬웨어 분류를 수행하였다. 실험 결과, 이진 분류에서는 특성 차원 축소기법이 성능 향상에 큰 영향을 미치지 않았으며, 다중 분류에서도 데이터세트의 특성 차원이 작을 경우에는 동일하였다. 그러나 학습데이터가 고차원의 특성을 가지는 상황에서 다중 분류를 시도했을 경우 LDA(Linear Discriminant Analysis)가 우수한 성능을 나타냈다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -3
• /
• pp.1420-1423
• /
• 2002

The classification and recognition of two-dimensional trademark patterns independently of their position, orientation, size and scale by proposing two feature vectors has been discussed. The paper presents experimentation on two feature vectors showing size- invariance and scale-invariance respectively. Both feature vectors are equally invariant to rotation as well. The feature extraction is based on local as well as global statistics of the image. These feature vectors have appealing mathematical simplicity and are versatile. The results so far have shown the best performance of the developed system based on these unique sets of feature. The goal has been achieved by segmenting the image using connected-component (nearest neighbours) algorithm. Second part of this work considers the possibility of using back propagation neural networks (BPN) for the learning and matching tasks, by simply feeding the feature vectosr. The effectiveness of the proposed feature vectors is tested with various trademarks, not used in learning phase.

• 한국멀티미디어학회논문지
• /
• 제15권2호
• /
• pp.195-203
• /
• 2012

본 논문은 다중센서 영상을 이용한 결정 융합 기반의 지상 표적 분류 알고리즘 및 특징 추출 기법을 제안한다. 표적의 인식률 향상을 위하여 가중 투표 방법을 적용함으로써 개별 분류기로부터 획득된 결과를 융합하였다. 또한 개별 센서 영상 내에 속한 표적을 분류하기 위해 CCD 영상으로부터 획득한 CM 영상의 밝기 차이와 FLIR 영상 내 표적의 윤곽선 정보 및 차량과 포탑의 너비 비율을 이용하여 스케일과 회전변화에 강인한 특징들을 추출하였다. 마지막으로 실험을 통하여 본 논문에서 제안한 지상 표적 분류 알고리즘과 특징 추출 기법에 대한 성능을 검증한다.

• 한국멀티미디어학회논문지
• /
• 제12권6호
• /
• pp.867-875
• /
• 2009

More and richer information sources and services are available on the web everyday. However, harmful information, such as adult content, is not appropriate for all users, notably children. Since internet is a worldwide open network, it has a limit to regulate users providing harmful contents through each countrie's national laws or systems. Additionally it is not a desirable way of developing a certain system-specific classification technology for harmful contents, because internet users can contact with them in diverse ways, for example, porn sites, harmful spams, or peer-to-peer networks, etc. Therefore, it is being emphasized to research and develop context-based core technologies for classifying harmful contents. In this paper, we propose an efficient text filter for blocking harmful texts of web documents using context-based technologies and examine which algorithms for feature selection, the process that select content terms, as features, can be useful for text categorization in all content term occurs in documents, are suitable for classifying harmful contents through implementation and experiment.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제51권3호
• /
• pp.122-128
• /
• 2002

In this paper, a robust pattern classification of breath sounds for the diagnosis of pulmonary dysfunction in noise environment is proposed. The feature parameter extraction method by highpass lifter algorithm and PM(projection measure) algorithm are used. 17 different groups of breath sounds are experimentally classified and investigated. The classification has been performed by 6 different types of combinations with proposed methods to evaluate the performances, such as ARC with EDM and LCC with EDM, WLCC with EDM, ARC with PM, LCC with PM, WLCC with PM. Furthermore, all feature parameters are extracted to 80th orders by 5th orders step, and all experiments are evaluated in increasing noise environments by degrees SNR 24dB to 0dB. As a results, WLCC which is derived from highpass lifter algorithm, is selected for the feature parameter extraction method. Pm is more robust than EDM in noisy environments to test and compare experimental results. WLCC with PM method(WLCC/PM) has a better performance in an increasing noise environment for diagnosis of pulmonary dysfunction.

• 제어로봇시스템학회논문지
• /
• 제15권8호
• /
• pp.778-785
• /
• 2009

This paper presents a Gabor texture feature extraction method for classification of discolored Metal pad images using GPU(Graphics Processing Unit). The proposed algorithm extracts the texture information using Gabor filters and constructs a pattern map using the extracted information. Finally, the golden pad images are classified by utilizing the feature vectors which are extracted from the constructed pattern map. In order to evaluate the performance of the Gabor texture feature extraction algorithm based on GPU, a sequential processing and parallel processing using OpenMP in CPU of this algorithm were adopted. Also, the proposed algorithm was implemented by using Global memory and Shared memory in GPU. The experimental results were demonstrated that the method using Shared memory in GPU provides the best performance. For evaluating the effectiveness of extracted Gabor texture features, an experimental validation has been conducted on a database of 20 Metal pad images and the experiment has shown no mis-classification.

• Journal of Computing Science and Engineering
• /
• 제10권2호
• /
• pp.58-67
• /
• 2016

With the large amount of complex network data that is increasingly available on the Web, link prediction has become a popular data-mining research field. The focus of this paper is on a link-prediction task that can be formulated as a binary classification problem in complex networks. To solve this link-prediction problem, a sparse-classification algorithm called "Truncated Kernel Projection Machine" that is based on empirical-feature selection is proposed. The proposed algorithm is a novel way to achieve a realization of sparse empirical-feature-based learning that is different from those of the regularized kernel-projection machines. The algorithm is more appealing than those of the previous outstanding learning machines since it can be computed efficiently, and it is also implemented easily and stably during the link-prediction task. The algorithm is applied here for link-prediction tasks in different complex networks, and an investigation of several classification algorithms was performed for comparison. The experimental results show that the proposed algorithm outperformed the compared algorithms in several key indices with a smaller number of test errors and greater stability.

• 한국컴퓨터정보학회논문지
• /
• 제28권4호
• /
• pp.1-11
• /
• 2023

본 논문은 순환 신경망 대신 합성곱 신경망을 사용하여 시계열 데이터 분류 성능을 분석한다. TSC(Time Series Community)에는 GAF(Gramian Angular Field), MTF(Markov Transition Field), RP(Recurrence Plot)와 같은 전통적인 시계열 데이터 이미지화 알고리즘들이 있다. 실험은 이미지화 알고리즘들에 필요한 하이퍼 파라미터들을 조정하면서 합성곱 신경망의 성능을 평가하는 방식으로 진행된다. UCR 아카이브의 GunPoint 데이터셋을 기준으로 성능을 평가했을 때, 본 논문에서 제안하는 STFT(Short Time Fourier Transform) 알고리즘이 최적화된 하이퍼 파라미터를 찾은 경우, 기존의 알고리즘들 대비 정확도가 높고, 동적으로 feature map 이미지의 크기도 조절가능하다는 장점이 있다. GAF 또한 98~99%의 높은 정확도를 보이지만, feature map 이미지의 크기를 동적으로 조절할 수 없어 크다는 단점이 존재한다.

• International Journal of Computer Science & Network Security
• /
• 제24권1호
• /
• pp.119-126
• /
• 2024

Disease caused by the coronavirus (COVID-19) is sweeping the globe. There are numerous methods for identifying this disease using a chest imaging. Computerized Tomography (CT) chest scans are used in this study to detect COVID-19 disease using a pretrain Convolutional Neural Network (CNN) ResNet50. This model is based on image dataset taken from two hospitals and used to identify Covid-19 illnesses. The pre-train CNN (ResNet50) architecture was used for feature extraction, and then fully connected layers were used for classification, yielding 97%, 96%, 96%, 96% for accuracy, precision, recall, and F1-score, respectively. When combining the feature extraction techniques with the Back Propagation Neural Network (BPNN), it produced accuracy, precision, recall, and F1-scores of 92.5%, 83%, 92%, and 87.3%. In our suggested approach, we use a preprocessing phase to improve accuracy. The image was enhanced using the Contrast Limited Adaptive Histogram Equalization (CLAHE) algorithm, which was followed by cropping the image before feature extraction with ResNet50. Finally, a fully connected layer was added for classification, with results of 99.1%, 98.7%, 99%, 98.8% in terms of accuracy, precision, recall, and F1-score.