http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
약물-표적 단백질 연관관계 예측모델을 위한 쌍 기반 뉴럴네트워크
이문환(Munhwan Lee),김응희(Eunghee Kim),김홍기(Hong-Gee Kim) 한국인지과학회 2017 인지과학 Vol.28 No.4
In-silico 기반의 약물-표적 단백질 연관관계 예측은 신약 탐색 단계에서 매우 중요하다. 그러나 기존의 예측모델은 입력 값이 고정적이며 표적 단백질의 특질 값이 가공된 데이터로 한정됨으로써 예측 모델의 확장성과 유연성이 부족하다. 본 논문에서는 약물-표적 단백질 연관관계를 예측하는 확장 가능한 형태의 머신러닝 모델을 소개한다. 확장 가능한 머신러닝 모델의 핵심 아이디어는 쌍기반의 뉴럴 네트워크로써, 약물과 단백질의 미가공 데이터를 사용하여 특질을 추출하고 특질 값을 각각의 뉴럴 네트워크 레이어에 입력한다. 이 방법은 추가적인 지식없이 자동적으로 약물과 단백질의 특질을 추출한다. 또한 쌍기반 레이어는 특질값을 풍부한 저차원의 벡터로 향상 시킴으로써 입력 값의 차이로 인한 편향 학습을 방지한다. PubChem BioAssay(PCBA) 데이터 셋에 기반한 5-폴드 교차 검증법을 통하여 제안한 모델의 성능을 평가했으며, 이전의 모델보다 우월한 성능을 보였다. Predicting compound-protein interactions in-silico is significant for the drug discovery. In this paper, we propose an scalable machine learning model to predict compound-protein interaction. The key idea of this scalable machine learning model is the architecture of pairwise neural network model and feature embedding method from the raw data, especially for protein. This method automatically extracts the features without additional knowledge of compound and protein. Also, the pairwise architecture elevate the expressiveness and compact dimension of feature by preventing biased learning from occurring due to the dimension and type of features. Through the 5-fold cross validation results on large scale database show that pairwise neural network improves the performance of predicting compound-protein interaction compared to previous prediction models.
김성수(Seongsoo Kim),김문환(Munhwan Kim),김순호(Soonho Kim) 한국자동차공학회 2002 한국자동차공학회 춘 추계 학술대회 논문집 Vol.2002 No.5_1
Engine-out HC emission were investigated during cold and hot start. The tests were conducted according to engine cooling temperatures which were controlled by simulated coolant temperatures of cold and hot start, on a I.5L, 4-cylinder, 16valve, multipoint-port-fuel-injection gasoline engine. Real time engine-out HC emissions were measured at exhaust ports and cylinder head using Fast Response Flame Ionization Detector (FRFID).<br/> Unburned hydrocarbons emitted at the cold coolant temperature were much higher than that of the hot coolant temperature. And the main source of the high HC emission was confirmed as misfire at cold coolant temperature. In addition, the effect of intake valve timing on engine-out HC emission was investigated. The results obtained indicate that optimized intake phasing provides the potential for start-up engine-out HC emission reduction.
김성수(Seongsoo Kim),김문환(Munhwan Kim),김순호(Soonho Kim) 한국자동차공학회 2003 한국자동차공학회 춘 추계 학술대회 논문집 Vol.- No.-
The effects of temperatures of catalytic converter, intake temperatures and fuel skip injections on engine-out HC emissions were investigated during cold and hot start. The tests were conducted according to engine cooling temperatures which were controlled by simulated coolant temperatures of cold and hot start, on a 1.5L, 4-cylinder, l6valve, multipoint-port-fuel-injection gasoline engine. Real time engine-out HC emissions were measured at a exhaust port and cylinder head using Fast Response Flame Ionization Detector (FRFID).<br/> Unburned hydrocarbons emitted at the cold coolant temperature were much higher than those of the hot coolant temperatures. And temperatures of catalytic converter and applications of fuel skip injections which can be controlled with coolant temperatures provide the high potential for start-up engine-out HC emissions reduction.<br/>
NP-HTM: 다중 코어 인메모리 데이터베이스에서 하드웨어 트랜잭셔널 메모리 분할 기법
강문환(Munhwan Kang),김형진(Hyeongjin Kim),마현국(Hyeonkuk Ma),장재우(Jaewoo Chang) 한국정보과학회 2018 정보과학회논문지 Vol.45 No.6
약 트랜잭셔널 메모리는 전통적인 병렬 프로그래밍 기법인 Lock을 대체하여 동시성 제어 패러다임을 크게 바꾸었다. 특히 HTM(Hardware Transactional Memory)은 하드웨어에 의해 지원되는 가장 뛰어난 기법이다. 그러나 기존 HTM 기법은 HTM 의 자원 제약성을 극복하지 못하는 문제점이 존재한다, 이러한 문제점을 해결하기 위해, 본 논문에서는 다중 코어 인 메모리 데이터베이스에서 대형 트랙잰션을 처리하기 위한 HTM 기반 트랜잭션 분할 기법을 제안한다. 첫째, 제안하는 기법은 트랜잭션이 자원 제약성 때문에 실패하면, 트랜잭션을 중첩 파티션 불럭으로 분할한다. 둘째, 제안하는 기법은 워크로드 특성을 반영하여 파티션 불럭의 최적의 길이를 계산하는 적응적 트랜잭션 분할 알고리즘을 사용한다. 마지막으로, STAMP 벤치마크를 사용한 실험적 성능 분석을 통해, 제안하는 기법이 기존의 트랜잭션 분할 기법인 Part-HTM에 비해 약 70%의 성능 향상을 보인다. Transactional Memory has greatly changed concurrency control paradigm by replacing locks, the conventional parallel programming mechanism. Especially, HTM(Hardware Transactional Memory) is the most promising scheme that is supported by hardware. However, the existing HTM techniques have a problem that they cannot overcome the resource limitations of HTM. To solve the problem, we propose a HTM-based transaction splitting technique to support large-sized transaction processing in multicore in-memory databases. First, the proposed technique can split a transaction into nested partition blocks when the transaction fails by resource limitation. Second, the proposed technique makes use of our adaptive split algorithm that computes the optimal size of partition blocks, according to the characteristic of a workload. Finally, through our experimental performance analysis using STAMP benchmark, the proposed technique shows about 70% better performance than the existing transaction splitting technique, i.e., Part-HTM.
Opportunistic Spatial Reuse in IEEE 802.15.3c Wireless Personal Area Networks
Sunggeun Jin,Munhwan Choi,Kyeongpyo Kim,Sunghyun Choi IEEE 2013 IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY Vol.62 No.2
<P>The IEEE 802.15.3c wireless personal area network (WPAN) standard is designed to support highly directional wireless communications at the 60-GHz frequency band. Highly directional communications are helpful to reduce interference; hence, great potentials are provided for aggregate network throughput improvement via spatial reuse (SR). However, in 802.15.3c WPANs, the SR is limited in use since the 802.15.3c standard does not specify an explicit procedure for the SR. We design a new and simple scheme exploiting the SR, thus improving aggregate network throughput. In the proposed scheme, the 802.15.3c devices measure and report the measured channel status. Then, a controlling device, which is called the piconet coordinator (PNC) in the 802.15.3c WPAN, schedules resources available for each device. The simulation results demonstrate that the proposed scheme contributes to significant aggregate network throughput improvement. The proposed scheme is so simple that it can be easily employed for practical 802.15.3c devices. In addition, it is guaranteed to work well without conflicting with existing 802.15.3c operations.</P>