http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
대량재해에 있어 미토콘드리아 DNA의 다형성을 이용한 개인식별
이숭덕,김기범,이윤성,최영태,신창호,이정빈 大韓法醫學會 1996 대한법의학회지 Vol.20 No.1
The mitochondrial DNA(mtDNA) is a small extranuclear DNA molecule that has been sequenced in man. the mtDNA has distinct characteristics over nuclear DNA. First, the evolutionary rate of nucleotide substitution appears to be larger compared with the nuclear DNA. Second, mtDNA is maternally inherited, so the determination of the various haplotype is unequivocal and no recombination has to be involved. Third, there is a 1.1 kb long control region, which shows severe polymorphism. These characteristics have been applied for several evolutionary study. Furthermore, they exist in cytoplasm with numerous copy number and their size, 16,569 bp is small relative to nuclear DNA, so they are more resistant to degradation or can be types in samples with only cytoplasm such as hair shaft without hair root. These presented the possibility of mtDNA as a tool in individual identification, especially when the nuclear DNA be unavailable. Actually, polymorphism in control region has been applied to individual identification for the skeletal remain found in Vietnam war. So we decided to apply the polymorphism of mtDNA in control region for the individual identification in case of mass disaster-"Downfall of Sampoong Department". Human remains from total 27 different individuals and 178 control persons were submitted for the individual identification. As the mtDNA comparison could be done through maternal lineage, 12 individual from 9 families from the control group were ruled out, because these families were consisted of father or were seeking for their father. Sequencing for the mtDNA control region was done using ABI automatic sequencer and DyeDeoxy Terminator Cycle Sequencing Kit. For convenience, sequencing reaction was done for two separate control region, region I from 16016-16401 in Anderson sequence, and regionⅡ from 048-388. First, sequences in regionⅡ of all samples and control persons were compared, and 20 samples were ruled out with no identical sequences in control group. Remaining 7 samples showed identical sequences with some of the control persons in three groups, and these identical pairs underwent mtDNA region I study. After region I study all samples except two dropped out in blood relationship. For the confirmation, STR and VNTR study such as MCT118, apoB, YNZ22, vWF, MBP, D21S11, TC-11 were done for remaining the two samples. The result denied the blood relationship between these two samples and any of the control group.
CT 방법을 이용한 질량감쇠계수 결정 및 자체감쇠 보정
이정빈,이준호,변종인,윤주용 (사)한국방사선산업학회 2018 방사선산업학회지 Vol.12 No.3
In this study, the mass attenuation coefficients for all five of the IAEA referencematerial samples (apparent density in a measuring bottle: 0.50~1.45 g·cm-3), including soil, milkpowder, hay and moss soil, were determined using the CT (Calibration Transmission) method. Acertified mixed gamma-ray sources including 241Am, 88Cd, 57Co, 113Sn, 85Sr and 137Cs were added tothe IAEA reference samples for the validation of present method. The self-attenuation correctionfactors for the gamma-ray energies of 59.5 keV, 88 keV, 122.1 keV, 391.7 keV, 514 keV and 661.7keV were determined and applied to the self-attenuation correction. As a result, the accuracy ofgamma-ray spectrometry for environmental samples used in this study was improved especiallyfor lower energy gamma-ray emitting radionuclides.
소스코드 주제를 이용한 인공신경망 기반 경고 분류 방법
이정빈,Lee, Jung-Been 한국정보처리학회 2020 정보처리학회논문지. 컴퓨터 및 통신시스템 Vol.9 No.11
자동화된 정적분석 도구는 소스 코드상에 잠재된 결함을 개발자들이 적은 노력으로 빠르게 찾을 수 있도록 도와준다. 하지만 이러한 정적분석 도구는 수정할 필요가 없는 오탐지 경고들을 무수하게 발생시킨다. 본 연구에서는 소스코드 블록의 토픽 모델을 이용한 인공신경망 기반의 경고 분류 기법을 제안한다. 소프트웨어 변경 관리 시스템으로부터 버그를 수정한 리비전들을 수집하고, 개발자들로부터 수정된 코드 블록들을 추출한다. 토픽 모델링을 이용하여 수집된 코드 블록의 토픽 분포 값을 구하고, 코드 블록의 리비전 간 경고들의 삭제 여부를 표현하는 이진데이터를 인공신경망의 입력 값과 출력 값으로 사용하여 심층 학습을 수행한다. 그 결과, 인공신경망 기반의 분류 모델이 높은 예측 성능으로 진성 또는 오탐지 경고를 분류하였다. Automatic Static Analysis Tools help developers to quickly find potential defects in source code with less effort. However, the tools reports a large number of false positive warnings which do not have to fix. In our study, we proposed an artificial neural network-based warning classification method using topic models of source code blocks. We collect revisions for fixing bugs from software change management (SCM) system and extract code blocks modified by developers. In deep learning stage, topic distribution values of the code blocks and the binary data that present the warning removal in the blocks are used as input and target data in an simple artificial neural network, respectively. In our experimental results, our warning classification model based on neural network shows very high performance to predict label of warnings such as true or false positive.