• 한국컴퓨터정보학회논문지
• /
• 제16권4호
• /
• pp.15-22
• /
• 2011

주성분 분석(PCA)은 데이터의 차원을 줄이면서 최대의 데이터 변이를 보존하는 기법으로 차원 축소나 특징 추출을 위해 널리 사용되고 있다. 하지만 PCA는 잡음에 민감하며 가우스 분포에 대하여만 유효하다는 단점이 있다. 잡음 민감성의 개선을 위해 다양한 방법이 제시되었고 그 중 퍼지 소속도를 이용한 반복적 최적화 기법인 RF-PCA2가 다른 방법에 비해 우수한 성능을 보였다. 하지만 RF-PCA2는 가우스 분포에만 사용할 수 있는 선형 알고리듬이라는 한계가 있다. 이 논문에서는 RF-PCA2와 커널 주성분 분석(kernel PCA, K-PCA)을 결합하여 가우스 분포 이외의 분포들도 다룰 수 있는 비선형 알고리듬인 improved robust kernel fuzzy PCA (RKF-PCA2)를 제안한다. RKF-PCA2는 RF-PCA2 알고리듬의 잡음 강건성과K-PCA의비선형성을 통해 기존알고리듬에 비해 잡음민감성이 적으며 가우스분포 한계를 효과적으로 극복할 수 있다. 이러한 사실은 실험 결과를 통해 확인할 수 있다.

• Structural Engineering and Mechanics
• /
• 제44권2호
• /
• pp.179-196
• /
• 2012

An eigenspace projection clustering method is proposed for structural damage detection by combining projection algorithm and fuzzy clustering technique. The integrated procedure includes data selection, data normalization, projection, damage feature extraction, and clustering algorithm to structural damage assessment. The frequency response functions (FRFs) of the healthy and the damaged structure are used as initial data, median values of the projections are considered as damage features, and the fuzzy c-means (FCM) algorithm are used to categorize these features. The performance of the proposed method has been validated using a three-story frame structure built and tested by Los Alamos National Laboratory, USA. Two projection algorithms, namely principal component analysis (PCA) and kernel principal component analysis (KPCA), are compared for better extraction of damage features, further six kinds of distances adopted in FCM process are studied and discussed. The illustrated results reveal that the distance selection depends on the distribution of features. For the optimal choice of projections, it is recommended that the Cosine distance is used for the PCA while the Seuclidean distance and the Cityblock distance suitably used for the KPCA. The PCA method is recommended when a large amount of data need to be processed due to its higher correct decisions and less computational costs.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권7호
• /
• pp.3076-3092
• /
• 2020

An IKPCA-ELM-based intrusion detection method is developed to address the problem of the low accuracy and slow speed of intrusion detection caused by redundancies and high dimensions of data in the network. First, in order to reduce the effects of uneven sample distribution and sample attribute differences on the extraction of KPCA features, the sample attribute mean and mean square error are introduced into the Gaussian radial basis function and polynomial kernel function respectively, and the two improved kernel functions are combined to construct a hybrid kernel function. Second, an improved particle swarm optimization (IPSO) algorithm is proposed to determine the optimal hybrid kernel function for improved kernel principal component analysis (IKPCA). Finally, IKPCA is conducted to complete feature extraction, and an extreme learning machine (ELM) is applied to classify common attack type detection. The experimental results demonstrate the effectiveness of the constructed hybrid kernel function. Compared with other intrusion detection methods, IKPCA-ELM not only ensures high accuracy rates, but also reduces the detection time and false alarm rate, especially reducing the false alarm rate of small sample attacks.

• 산업공학
• /
• 제24권1호
• /
• pp.8-14
• /
• 2011

Principal Component Analysis (PCA) reduces the dimensionality of the process by creating a new set of variables, Principal components (PCs), which attempt to reflect the true underlying process dimension. However, for highly nonlinear processes, this form of monitoring may not be efficient since the process dimensionality can't be represented by a small number of PCs. Examples include the process of semiconductors, pharmaceuticals and chemicals. Nonlinear correlated process variables can be reduced to a set of nonlinear principal components, through the application of Kernel Principal Component Analysis (KPCA). Support Vector Data Description (SVDD) which has roots in a supervised learning theory is a training algorithm based on structural risk minimization. Its control limit does not depend on the distribution, but adapts to the real data. So, in this paper proposes a non-linear process monitoring technique based on supervised learning methods and KPCA. Through simulated examples, it has been shown that the proposed monitoring chart is more effective than $T^2$ chart for nonlinear processes.

• Journal of Information Processing Systems
• /
• 제17권1호
• /
• pp.178-190
• /
• 2021

Alzheimer's disease (AD) is an insidious and degenerative neurological disease. It is a new topic for AD patients to use magnetic resonance imaging (MRI) and computer technology and is gradually explored at present. Preprocessing and correlation analysis on MRI data are firstly made in this paper. Then kernel principal component analysis (KPCA) is used to extract features of brain gray matter images. Finally supervised classification schemes such as AdaBoost algorithm and support vector machine algorithm are used to classify the above features. Experimental results by means of AD program Alzheimer's Disease Neuroimaging Initiative (ADNI) database which contains brain structural MRI (sMRI) of 116 AD patients, 116 patients with mild cognitive impairment, and 117 normal controls show that the proposed method can effectively assist the diagnosis and analysis of AD. Compared with principal component analysis (PCA) method, all classification results on KPCA are improved by 2%-6% among which the best result can reach 84%. It indicates that KPCA algorithm for feature extraction is more abundant and complete than PCA.

• 한국작물학회지
• /
• 제50권spc1호
• /
• pp.161-166
• /
• 2005

작물과학원에서 육성한 찰옥수수 자식계통을 대상으로 농업적 형질 및 품질관련 특성 등을 분석하여 계통군화 시켜 교배 모, 부본의 기초 자료로 이용하고 품질육종의 효율성을 높이기 위한 연구결과를 요약하면 다음과 같다. 1. 작물과학원에서 육성한 찰옥수수 64개 자식계통에 대해서 옥수수 낱알형태를 조사한 결과 백립중은 $11.7\~37.3g$,과 피두께 $11\~77mm$, 립장 $5.8\~9.6mm$, 립폭 $6.5\~10.0mm$, 립두께 $4.1\~6.8mm$의 분포를 나타내었다. 2. 이화학적 특성 분석에 있어서는 단백질 함량은 $8.7\~15.8\%$, 지방 $2.3\~5.8\%$, 유리당 $1.1\~11.0\%$, 아밀로펙틴 $78.5\~93.8\%$의 다양한 변이분포를 보였다. 3. 아밀로그램을 분석한 결과 초기호화온도는 $64.5\~79.1^{\circ}C$, 최고점도 $14.8\~221.8$ RVU, 최저점도 $12.2\~109.6$ RVU, 최종점도 $20.1\~212.6$ RVU였으며 강하점도는 $2.0\~l12.2$ RVU, 치반점도 $-81.3\~85.3$ RVU, 응집점도 $7.9\~105.0$ RVU로 나타났다. 4. Texture 분석을 한 결과 껌성 $91\~383$, 경도 $181\~394$, 씹힘성 $73\~370$의 범위에 속하였다. 5. 64개 자식계통을 대상으로 단백질 등 14개 형질을 이용하여 주성분 분석을 한 결과 제 1주성분은 단백질과 지방 함량이 높고 경도가 높은 것으로 나타나 주로 저식미 계통군으로 분류되며 백립중과 립장이 부의 값이어서 소립의 형태를 보였으며 제2주성분은 립장, 립폭, 백립중 등이 정의 값을 나타내어 대립의 계통군들이 포함되며 경도와는 부의 상관을 보여 식미가 양호한 계통군으로 분류되었다. 6. 제1주성분과 제2주성분을 이용하여 군집분석을 한 결과 8개 자식계통군으로 구분할 수 있었으며 scatter diagram에서 제1주성분이 작아지고 제2주성분이 커지는 좌측 상단부에 분포한 VII, VIII군에 속한 계통들이 주로 식미가 높으며 조숙종보다는 중, 만숙종들이 대부분 이였다.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2006년도 추계학술대회논문집
• /
• pp.401-406
• /
• 2006

This paper presents the fault diagnosis of induction motor based on support vector machine(SVMs). SVMs are well known as intelligent classifier with strong generalization ability. Application SVMs using kernel function is widely used for multi-class classification procedure. In this paper, the algorithm of SVMs will be combined with feature extraction and reduction using component analysis such as independent component analysis, principal component analysis and their kernel(KICA and KPCA). According to the result, component analysis is very useful to extract the useful features and to reduce the dimensionality of features so that the classification procedure in SVM can perform well. Moreover, this method is used to induction motor for faults detection based on vibration and current signals. The results show that this method can well classify and separate each condition of faults in induction motor based on experimental work.

• 한국해양공학회:학술대회논문집
• /
• 한국해양공학회 2006년 창립20주년기념 정기학술대회 및 국제워크샵
• /
• pp.46-53
• /
• 2006

This paper presents the fault diagnosis of induction motor based on support vector machine (SVMs). SVMs are well known as intelligent classifier with strong generalization ability. Application SVMs using kernel function is widely used for multi-class classification procedure. In this paper, the algorithm of SVMs will be combined with feature extraction and reduction using component analysis such as independent component analysis, principal component analysis and their kernel (KICA and KPCA). According to the result, component analysis is very useful to extract the useful features and to reduce the dimensionality of features so that the classification procedure in SVM can perform well. Moreover, this method is used to induction motor for faults detection based on vibration and current signals. The results show that this method can well classify and separate each condition of faults in induction motor based on experimental work.

• 대한음성학회지:말소리
• /
• 제66호
• /
• pp.105-116
• /
• 2008

In this research, we propose a speaker identification system using a kernel method which is expected to model the non-linearity of speech features well. We have been using principal component analysis (PCA) successfully, and extended to kernel PCA, which is used for many pattern recognition tasks such as face recognition. However, we cannot use kernel PCA for speaker identification directly because the storage required for the kernel matrix grows quadratically, and the computational cost grows linearly (computing eigenvector of $l{\times}l$ matrix) with the number of training vectors I. Therefore, we use greedy kernel PCA which can approximate kernel PCA with small representation error. In the experiments, we compare the accuracy of the greedy kernel PCA with the baseline Gaussian mixture models using MFCCs and PCA. As the results with limited enrollment data show, the greedy kernel PCA outperforms conventional methods.

• 한국지능시스템학회논문지
• /
• 제18권3호
• /
• pp.297-305
• /
• 2008

최근들어, 커널 기법(kernel method)은 패턴 분류, 함수 근사 및 비정상 상태 탐지 등의 분야에서 상당한 관심을 끌고 있다. 특히, 서포트 벡터 머신(support vector machine)이나 커널 주성분 분석(kernel principal component analysis) 등의 방법론에서 커널의 역할은 매우 중요한데, 이는 고전적인 선형 머신이 비선형성을 효과적으로 다룰 수 있도록 일반화 해줄 수 있기 때문이다. 본 논문에서는 커널 기반 가우시안 프로세스(gaussian process) 함수근사 기법과 서포트 벡터 학습을 이용하여 레이더와 강우계의 관측 데이터를 융합하는 문제를 고려한다. 그리고, 국내의 강원, 경북 및 충북에 걸쳐있는 지역에 대한 레이더 자료 및 강우계 자료를 대상으로 하여 본 논문에서 고려하는 방법론들에 의해 데이터 융합을 수행한 결과를 제시하고, 성능비교를 수행한다.