• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1997년도 한국자동제어학술회의논문집; 한국전력공사 서울연수원; 17-18 Oct. 1997
• /
• pp.664-667
• /
• 1997

For higher component density per chip, it is necessary to identify and control the semiconductor manufacturing process more stringently. Recently, neural networks have been identified as one of the most promising techniques for modeling and control of complicated processes such as plasma etching process. Since wafer states after each run using identical recipe may differ from each other, conventional neural network models utilizing input factors only cannot represent the actual state of process and equipment. In this paper, in addition to the input factors of the recipe, real-time tool data are utilized for modeling of 64M DRAM s-poly plasma etching process to reflect the actual state of process and equipment. For real-time tool data, we collect optical emission spectroscopy (OES) data. Through principal component analysis (PCA), we extract principal components from entire OES data. And then these principal components are included to input parameters of neural network model. Finally neural network model is trained using feed forward error back propagation (FFEBP) algorithm. As a results, simulation results exhibit good wafer state prediction capability after plasma etching process.

• International Journal of Oral Biology
• /
• 제43권2호
• /
• pp.101-111
• /
• 2018

Understanding the classification of malocclusion is a crucial issue in Orthodontics. It can also help us to diagnose, treat, and understand malocclusion to establish a standard for definite class of patients. Principal component analysis (PCA) and k-means algorithms have been emerging as data analytic methods for cephalometric measurements, due to their intuitive concepts and application potentials. This study analyzed the macro- and meso-scale classification structure and feature basis vectors of 1020 (415 male, 605 female; mean age, 25 years) orthodontic patients using statistical preprocessing, PCA, random matrix theory (RMT) and k-means algorithms. RMT results show that 7 principal components (PCs) are significant standard in the extraction of features. Using k-means algorithms, 3 and 6 clusters were identified and the axes of PC1~3 were determined to be significant for patient classification. Macro-scale classification denotes skeletal Class I, II, III and PC1 means anteroposterior discrepancy of the maxilla and mandible and mandibular position. PC2 and PC3 means vertical pattern and maxillary position respectively; they played significant roles in the meso-scale classification. In conclusion, the typical patient profile (TPP) of each class showed that the data-based classification corresponds with the clinical classification of orthodontic patients. This data-based study can provide insight into the development of new diagnostic classifications.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2015년도 제46회 하계학술대회
• /
• pp.1345-1346
• /
• 2015

본 논문에서는 보행자 및 배경 이미지로부터 HOG-PCA 특징을 추출하고 다항식 기반 RBFNNs(Radial Basis Function Neural Network) 패턴분류기과 최적화 알고리즘을 이용하여 보행자를 검출하는 시스템 설계를 제안한다. 입력 영상으로부터 보행자를 검출하기 위해 전처리 과정에서 HOG(Histogram of oriented gradient) 알고리즘을 통해 특징을 추출한다. 추출된 특징은 고차원이므로 패턴분류기 분류 시 많은 연산과 처리속도가 따른다. 이를 개선하고자 PCA (Principal Components Analysis)을 사용하여 저차원으로의 차원 축소한다. 본 논문에서 제안하는 분류기는 pRBFNNs 패턴분류기의 효율적인 학습을 위해 최적화 알고리즘인 PSO(Particle Swarm Optimization)을 사용하여 구조 및 파라미터를 최적화시켜 모델의 성능을 향상시킨다. 사용된 데이터로는 보행자 검출에 널리 사용되는 INRIA2005_person data set에서 보행자와 배경 영상을 각각 1200장을 학습 데이터, 검증 데이터로 구성하여 분류기를 설계하고 테스트 이미지를 설계된 최적의 분류기를 이용하여 보행자를 검출하고 검출률을 확인한다.

• Preventive Nutrition and Food Science
• /
• 제21권2호
• /
• pp.138-146
• /
• 2016

In this study, we report the optimal extraction conditions for obtaining organosulfur compounds, such as cycloalliin, from garlic by using principal component analysis (PCA). Extraction variables including temperature ($40{\sim}80^{\circ}C$), time (0.5~12 h), and pH (4~12) were investigated for the highest cycloalliin yields. The cycloalliin yield (5.5 mmol/mL) at pH 10 was enhanced by ~40% relative to those (~3.9 mmol/mL) at pH 4 and pH 6. The cycloalliin level at $80^{\circ}C$ showed the highest yield among the tested temperatures (5.05 mmol/mL). Prolonged extraction times also increased cycloalliin yield; the yield after 12 h was enhanced ~2-fold (4 mmol/mL) compared to the control. Isoalliin and cycloalliin levels were inversely correlated, whereas a direct correlation between polyphenol and cycloalliin levels was observed. In storage for 30 days, garlic stored at $60^{\circ}C$ (11 mmol/mL) showed higher levels of cycloalliin and polyphenols than those at $40^{\circ}C$, with the maximum cycloalliin level (13 mmol/mL) on day 15. Based on the PCA analysis, the isoalliin level depended on the extraction time, while cycloalliin amounts were influenced not only by extraction time, but also by pH and temperature. Taken together, extraction of garlic at $80^{\circ}C$, with an incubation time of 12 h, at pH 10 afforded the maximum yield of cycloalliin.

• 한국대기환경학회지
• /
• 제18권4호
• /
• pp.317-325
• /
• 2002

An air pollution monitoring station has been operated at Tokchok Island since April 1999 to characterize the background atmosphere in the vicinity of the Yellow Sea. In this study, eight chemical species in PM$_{2.5}$ and three gaseous species were analyzed. A total of 53 samples were collected for the analysis of PM$_{2.5}$ and gaseous species from April, 1999 to April, 2001. The overall mean mass concentration of PM$_{2.5}$ was 20.8 $\mu\textrm{g}$/㎥ and the eight soluble ionic species accounted for about 46.8% of PM$_{2.5}$ mass. Approximately 80% of samples appeared to experience the chloride loss effect. Air pollutant sources of PM$_{2.5}$ measured at Tokchok Island were qualitatively identified by the principal component analysis. It was found that five principal components are secondary aerosol, soil, incineration, phase change of nitrate, and ocean.and ocean.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권8호
• /
• pp.3865-3883
• /
• 2016

Infrastructure as a Service (IaaS) encapsulates computer hardware into a large amount of virtual and manageable instances mainly in the form of virtual machine (VM), and provides rental service for users. Currently, VM anomaly incidents occasionally occur, which leads to performance issues and even downtime. This paper aims at detecting anomalous VMs based on performance metrics data of VMs. Due to the dynamic nature and increasing scale of IaaS, detecting anomalous VMs from voluminous correlated and non-Gaussian monitored performance data is a challenging task. This paper designs an anomaly detection framework to solve this challenge. First, it collects 53 performance metrics to reflect the running state of each VM. The collected performance metrics are testified not to follow the Gaussian distribution. Then, it employs independent components analysis (ICA) instead of principal component analysis (PCA) to extract independent components from collected non-Gaussian performance metric data. For anomaly detection, it employs multi-class Bayesian classification to determine the current state of each VM. To evaluate the performance of the designed detection framework, four types of anomalies are separately or jointly injected into randomly selected VMs in a campus-wide testbed. The experimental results show that ICA-based detection mechanism outperforms PCA-based and LDA-based detection mechanisms in terms of sensitivity and specificity.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2007년도 추계학술대회논문집
• /
• pp.1430-1436
• /
• 2007

A principal components analysis (PCA) of the median head-related impulse responses (HRIRs) in the CIPIC HRTF database reveals that the individual HRIRs can be adequately reconstructed by a linear combination of 12 orthonormal basis functions. These basis functions can be used generally to model arbitrary HRIRs, which are not included in the process to obtain the basis functions. To clarify whether these basis functions can be used to model other set of arbitrary HRIRs, an numerical error analysis for modeling and a series of subjective listening tests were carried out using the measured and modeled HRIRs. The results showed that the set of individual HRIRs, which were measured in our lab using different measurement conditions, techniques, and source positions, can be well modeled with reasonable accuracy. Furthermore, all subjects reported not only the accurate vertical perception but also the front-back discrimination with the modeled HRIRs based on 12 basis functions. However, as less basis functions were used for HRIR modeling, the modeling accuracy and localization performance deteriorated.

• 한국소음진동공학회논문집
• /
• 제18권4호
• /
• pp.448-457
• /
• 2008

A principal components analysis (PCA) of the median-plane head-related impulse responses (HRIRs) in the CIPIC HRTF database reveals that the individual HRIRs in the median plane can be adequately reconstructed by a linear combination of 12 orthonormal basis functions. These basis functions can be used to model arbitrary median-plane HRIRs, which are not included in the process to obtain the basis functions. Memory size can be reduced up to 5-fold depending on the number of HRIRs to be modeled. To clarify whether these basis functions can be used to model other set of arbitrary median plane HRIRs, a numerical error analysis for modeling and a series of subjective listening tests were carried out using the measured and modeled HRIRs. The results showed that the set of individual HRIRs in the median plane, which were measured in our lab using different measurement conditions, techniques, and source positions, can be modeled with reasonable accuracy. All subjects, involved in the subjective listening test, reported not only the accurate vertical perception but also the front-back discrimination with the modeled HRIRs based on 12 basis functions.

• 한국환경생태학회지
• /
• 제29권2호
• /
• pp.145-161
• /
• 2015

This research was conducted in order to understand the hybridization between Quercus aliena Blume and Q. serrata Murray in Korea which show wide range of morphological variations within species and interspecific variations of diverse overlapping characteristics caused by hybridization. Morphological analysis (principal components analysis; PCA) of 116 individuals representing two species and their intermediates were performed. As a result, two species were clearly distinguished in terms of morphology, and intermediate morpho-types assumed to be hybrids between the two species were mostly located in the middle of each parent species in the plot of the principal components analysis. There was a clear distinction between two species in trichome distribution pattern which is an important diagnostic character in taxonomy of genus Quercus, whereas intermediate morpho-types showed intermediate state between two species' trichome distributions. Forty-two individuals representing two species and their intermediates were examined for leaf flavonoid constituents. Twenty-three flavonoid compounds were isolated and identified: They were glycosylated derivatives of flavonols, kaempferol, quercetin, isorhamnetin and myricetin. The flavonoid constituents of Q. aliena were five glycosylated derivatives: kaempferol 3-O-galactoside, kaempferol 3-O-glucoside, quercetin 3-O-galactoside, quercetin 3-O-glucoside, and Isorhamnetin 3-O-glucoside. The flavonoid constituents of Q. serrata had 20 diverse flavonol compounds including five flavonoid compounds found in Q. aliena. It was found that there is a clear difference in flavonoid constituents of Q. aliena and Q. serrata. Flavonoid chemistry is very useful in recognizing each species and putative hybrids. The flavonoid constituents of intermediates were a mixture of the two species' constituents and they generally showed similar characteristics to morpho-types. The hybrids between Q. aliena and Q. serrata showed morphologically and chemically diverse characteristics and it is assumed that there are frequent interspecific hybridization and introgression.

• 한국음향학회지
• /
• 제22권7호
• /
• pp.539-544
• /
• 2003

본 논문은MLLR (Maximum Likelihood Linear Regression)를 화자 적응시 과적응 방지를 위해 트리 구조에서 HHM 파라메타의 변환을 결정하는 점유 문턱값 (occupation threshold)의 영향을 감소하는 방법에 대해 기술한다. 데이터의 특징을 잘 나타내는 주성분 분석과 독립성분 분석을 통해 모델 혼합성분의 상관관계를 줄이고 상대적으로 데이터의 분포가 적은 축을 삭제함으로써 적은 적응데이터에 의한 과적응의 영향을 감소시켰다. 점유 문턱값을 작게 설정함으로써 변환함수의 수를 증가시켰을 경우, 기존의 MLLR 알고리즘은 과적응에 의해 화자 독립 모델보다 낮은 인식률을 나타내는 반면, 제안한 MLLR알고리즘은 화자 독립 모델의 성능에 비해 평균 2%이상 인식율 향상을 나타내었다.