• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2012년도 추계학술대회
• /
• pp.392-393
• /
• 2012

본 논문은 군집과 위키피디아(Wikipedia)를 이용하여 문서를 군집하는 새로운 방법을 제안한다. 제안된 방법은 비음수행렬분해를 이용하여 군집을 대표할 수 있는 군집 주제(topic)의 개념을 잘 표현할 수 있으며, 위키피디아의 동음이의어를 사용함으로써 문서와 군집 간의 의미관계를 고려하지 않는 용어집합(bag-of-words) 문제를 해결할 수 있다. 실험결과 제안방법을 적용한 문서군집방법이 다른 문서군집 방법에 비하여 좋은 성능을 보인다.

• 한국정보통신학회논문지
• /
• 제16권11호
• /
• pp.2374-2381
• /
• 2012

본 논문은 퍼지와 의미특징을 이용한 새로운 스니핏 추출의 성능향상 방법을 제안한다. 제안방법은 문장집합의 의미특징을 이용하여 대표문장을 생성하고, 대표문장과 문장집합의 퍼지관계를 이용함으로써 질의를 잘 나타내는 스니핏을 추출한다. 또한 의사연관 피드백을 이용하여 질의를 확장함으로써 사용자의 의도가 의미적으로 더 잘 포함되는 스니핏을 추출할 수 있다. 실험결과 제안방법이 다른 방법에 비해서 스니핏 추출에 더 좋은 성능을 보인다.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2004년도 봄 학술발표논문집 Vol.31 No.1 (B)
• /
• pp.688-690
• /
• 2004

최근 전자화된 문서 영상을 효율적으로 관리하고 검색하기 위한 문서구조분석 방법과 문서의 자동 분류에 관한 많은 연구가 발표되고 있다. 본 논문에서는 NMF(non-negative matrix factorization) 알고리즘을 사용하여 폰트를 자동으로 분류하는 방법을 제안한다. 제안된 방법은 폰트의 구분 특징들이 공간적으로 국부성을 가지는 부분으로 표현될 수 있다는 가정을 바탕으로, 전체의 폰트 이미지들로부터 각 폰트들의 구분 특징인 부분을 학습하고, 학습된 부분들을 특징으로 사용하여 폰트를 분류하는 방법이다. 학습된 폰트의 특징들은 계층적 군집화 알고리즘을 이용하여 템플릿을 생성하고, 테스트 패턴을 분류하기 위하여 템플릿 패턴과의 EMD(earth mover's distance)를 사용한다. 실험결과에서 폰트 이미지들의 공간적으로 국부적인 특징들이 조사되고, 그 특징들의 폰트 식별을 위한 적절성을 보였다. 제안된 방법이 기존의 문자인식. 문서 검색 시스템들의 전처리기로 사용되면. 그 시스템들의 성능을 향상시킬 것으로 기대된다.

• The Korean Journal of Physiology and Pharmacology
• /
• 제23권4호
• /
• pp.237-249
• /
• 2019

Confirming the direct link between neural circuit activity and animal behavior has been a principal aim of neuroscience. The genetically encoded calcium indicator (GECI), which binds to calcium ions and emits fluorescence visualizing intracellular calcium concentration, enables detection of in vivo neuronal firing activity. Various GECIs have been developed and can be chosen for diverse purposes. These GECI-based signals can be acquired by several tools including two-photon microscopy and microendoscopy for precise or wide imaging at cellular to synaptic levels. In addition, the images from GECI signals can be analyzed with open source codes including constrained non-negative matrix factorization for endoscopy data (CNMF_E) and miniscope 1-photon-based calcium imaging signal extraction pipeline (MIN1PIPE), and considering parameters of the imaged brain regions (e.g., diameter or shape of soma or the resolution of recorded images), the real-time activity of each cell can be acquired and linked with animal behaviors. As a result, GECI signal analysis can be a powerful tool for revealing the functions of neuronal circuits related to specific behaviors.

• 한국컴퓨터정보학회논문지
• /
• 제28권8호
• /
• pp.39-48
• /
• 2023

본 연구는 의류의 디자인 간 치수의 불일치와 비표준화로 인해 온라인 구매 시 발생하는 치수 선택의 오류 문제를 해결할 수 있는 방안을 제시하기 위해 수행되었다. 본 논문은 구매자에게 개인화된 치수를 제시할 수 있는 기계 학습 기반 추천 시스템의 구현 방안을 다루고 있다. 온라인 상거래로부터 발생된 구매 데이터를 사용하여 비음수 행렬 분해(NMF), 특이값 행렬 분해(SVD), k-최근접 이웃(KNN), 공동 클러스터링(Co-Clustering) 등 여러 검증된 협업 필터링 알고리즘을 훈련하였고, 이들 간에 성능을 비교하였다. 연구 결과, 비음수 행렬 분해 (NMF) 알고리즘이 다른 알고리즘들보다 뛰어난 성능을 보임을 확인할 수 있었다. 동일한 계정을 사용하는 여러 구매자가 포함되는 구매 데이터의 특성에도 불구하고, 제안 모형은 충분한 정확도를 보였다. 본 연구의 결과는 치수 선택의 오류로 인한 반품률을 감소하고 전자상거래 플랫폼에서의 고객 경험을 향상시키는 데 기여할 것으로 기대된다.

• 한국진공학회:학술대회논문집
• /
• 한국진공학회 2012년도 제43회 하계 정기 학술대회 초록집
• /
• pp.151-151
• /
• 2012

A new plasma process, i.e., the combination of PIII&D and HIPIMS, was developed to implant non-gaseous ions into materials surface. HIPIMS is a special mode of operation of pulsed-DC magnetron sputtering, in which high pulsed DC power exceeding ~1 kW/$cm^2$ of its peak power density is applied to the magnetron sputtering target while the average power density remains manageable to the cooling capacity of the equipment by using a very small duty ratio of operation. Due to the high peak power density applied to the sputtering target, a large fraction of sputtered atoms is ionized. If the negative high voltage pulse applied to the sample stage in PIII&D system is synchronized with the pulsed plasma of sputtered target material by HIPIMS operation, the implantation of non-gaseous ions can be successfully accomplished. The new process has great advantage that thin film deposition and non-gaseous ion implantation along with in-situ film modification can be achieved in a single plasma chamber. Even broader application areas of PIII&D technology are believed to be envisaged by this newly developed process. In one application of non-gaseous plasma immersion ion implantation, Ge ions were implanted into SiO2 thin film at 60 keV to form Ge quantum dots embedded in SiO2 dielectric material. The crystalline Ge quantum dots were shown to be 5~10 nm in size and well dispersed in SiO2 matrix. In another application, Ag ions were implanted into SS-304 substrate to endow the anti-microbial property of the surface. Yet another bio-application was Mg ion implantation into Ti to improve its osteointegration property for bone implants. Catalyst is another promising application field of nongaseous plasma immersion ion implantation because ion implantation results in atomically dispersed catalytic agents with high surface to volume ratio. Pt ions were implanted into the surface of Al2O3 catalytic supporter and its H2 generation property was measured for DME reforming catalyst. In this talk, a newly developed, non-gaseous plasma immersion ion implantation technique and its applications would be shown and discussed.

• 대한임상검사과학회지
• /
• 제52권3호
• /
• pp.221-228
• /
• 2020

Ochrobactrum anthropi는 oxidase를 생산하는 비발효산소성 그람음성 막대균으로 외관이 비슷하고 oxidase가 양성인 비발효 세균과 혼합배양 시 구분이 힘들고 생화학적 동정 장비로는 정확한 동정에 한계가 있다. 따라서 본 연구에서는 생화학적 검사 방법으로 동정이 힘든 세균동정의 Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry Platform (MALDI-TOF) 법의 유용성을 제시하고자 하였다. MicroScan을 이용해 검사했던 O. anthropi 5례를 분석한 결과, 최종보고까지 6.2일이 소요되었으며 E. coli의 3.0일에 비해 3.5일이 더 소요되었다. 5번 환자 고름 검체는 Achromobater xylosoxidans와 혼합감염으로 여러 번의 계대 배양과 재검사로 인해 11.3일이 소요되었는데, MALDI-TOF법으로 검사한 경우 한 번에 동정되었다. 4명의 환자는 기저질환이 있는 60세 이상이었고 기회감염과 원내감염의 가능성을 배제할 수 없었으며, 그 중 92일 만에 채취된 검체는 imipenem과 meropenem에 내성이었다. 따라서 O. anthropi처럼 동정이 까다로운 세균은 신속하고 적절한 환자 치료를 위해 MALDI-TOF법을 이용한 검사가 매우 유용할 것으로 사료된다.

• Journal of Periodontal and Implant Science
• /
• 제48권4호
• /
• pp.251-260
• /
• 2018

Purpose: The aim of this retrospective cross-sectional study was to evaluate whether salivary findings of active matrix-metalloproteinase 8 (aMMP-8) chairside (point of care; POC) tests were associated with periodontal risk assessment parameters in patients receiving supportive periodontal therapy (SPT). Methods: A total of 125 patients receiving regular SPT were included, and their records were examined. The following inclusion criteria were used: a diagnosis of chronic periodontitis, at least 1 non-surgical periodontal treatment (scaling and root planning) with following regular SPT (minimum once a year), at least 6 remaining teeth, and clinical and aMMP-8 findings that were obtained at the same appointment. In addition to anamnestic factors (e.g., smoking and diabetes), oral hygiene indices (modified sulcus bleeding index [mSBI] and approximal plaque index), periodontal probing depth simultaneously with bleeding on probing, and dental findings (number of decayed, missing, and filled teeth) were recorded. Salivary aMMP-8 levels were tested using a commercial POC test system (Periomarker, Hager & Werken, Duisburg, Germany). Statistical analysis was performed using the t-test, Mann-Whitney U test, Fisher's exact test, and ${\chi}^2$ test, as appropriate (P<0.05). Results: Only the mSBI was significantly associated with positive salivary aMMP-8 findings (aMMP-8 positive: $27.8%{\pm}20.9%$ vs. aMMP-8 negative: $18.0%{\pm}14.5%$; P=0.017). No significant associations were found between aMMP-8 and smoking, diabetes, periodontal parameters, or parameters related to the maintenance interval (P>0.05). Conclusions: Salivary aMMP-8 chairside findings were not associated with common parameters used for periodontal risk assessment in patients receiving SPT. The diagnostic benefit of POC salivary aMMP-8 testing in risk assessment and maintenance interval adjustment during SPT remains unclear.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제4권1호
• /
• pp.25-44
• /
• 2010

With the increasing popularity of mobile devices, it has become necessary to protect private information and content in these devices. Face recognition has been favored over conventional passwords or security keys, because it can be easily implemented using a built-in camera, while providing user convenience. However, because mobile devices can be used both indoors and outdoors, there can be many illumination changes, which can reduce the accuracy of face recognition. Therefore, we propose a new face recognition method on a mobile device robust to illumination variations. This research makes the following four original contributions. First, we compared the performance of face recognition with illumination variations on mobile devices for several illumination normalization procedures suitable for mobile devices with low processing power. These include the Retinex filter, histogram equalization and histogram stretching. Second, we compared the performance for global and local methods of face recognition such as PCA (Principal Component Analysis), LNMF (Local Non-negative Matrix Factorization) and LBP (Local Binary Pattern) using an integer-based kernel suitable for mobile devices having low processing power. Third, the characteristics of each method according to the illumination va iations are analyzed. Fourth, we use two matching scores for several methods of illumination normalization, Retinex and histogram stretching, which show the best and $2^{nd}$ best performances, respectively. These are used as the inputs of an SVM (Support Vector Machine) classifier, which can increase the accuracy of face recognition. Experimental results with two databases (data collected by a mobile device and the AR database) showed that the accuracy of face recognition achieved by the proposed method was superior to that of other methods.

• Mass Spectrometry Letters
• /
• 제8권4호
• /
• pp.109-113
• /
• 2017

Single analytical procedure including extraction, liquid chromatography, and mass spectrometric analysis was evaluated for the simultaneous measurement of lysophospholipids (LPLs). LPLs, particularly, lysophosphatidic acids (LPA) and sphingosine 1-phosphate (S1P) are lipid messengers ubiquitously found in various biological matrix. The molecular species mediate important physiological roles in association with many diseases (e.g. cancer, inflammation, and neurodegenerative disease), which emphasize the significance of the simple and reliable analytical method for biomarker discovery and molecular mechanistic understanding. Thus, we developed analytical method mainly focusing on, but not limited by those lipid species S1P and LPA using reverse phase liquid chromatography-tandem mass spectrometry (RPLC-ESI-MS-MS). Extraction method was modified based on Folch method with optimally minimal level of ionization additive (ammonium formate 10 mM and formic acid). Reverse-phase liquid-chromatography was applied for chromatographical separation in combination with negative ionization mode electrospray-coupled Orbitrap mass spectrometry. The method validation was performed on human blood plasma in a non-targeted lipid profiling manner with full-scan MS mode and data-dependent MS/MS. The proposed method presented good inter-assay precision for primary targets, S1P and LPA. Subsequent analysis of other types of LPLs identified a broad range of lysophosphatidylcholines (LPCs) and lysophosphatidyl-ethanolamines (LPEs).