http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
洪昌基,金潞經,金丁龍 中央醫學社 1941 中央醫學 Vol.10 No.1
The authors investigated the effects of acid stable digestivee enzyme preparation "Santorase" as a digestant on the in digestive symptoms in patients with various gastrointestinal disorders, liver: diseases, and on digestive symptoms following long term administrations of antibiotics and anti-tuberculosis agents, who visited to the Department of Internal Medicine, Seoul National University Hospital. The cases dealing with this report were shown as in table 1 to 4. The results were summarized as follows 1) The Clinical effects of Santorase on the digestive action were good without being influenced by the gastric acidity. 2) Combined administrations of this enzyme preparation with the other digestives orally promoted the digestive action and relieved the symptoms in the patients with gastrointestinal disorders and hepatic disorders. 3) Combined administrations of this enzyme preparation with. Antibio2tics and anti-tuberculosis agents relieved the side effects of these agents on the gastrointestinal system. 4) The valuability and possibility use of the enzyme preparation with the other digestives was discussed also.
Numerical Simulation of Conductance through Quantum Point Contacts with an Embedded Impurity
홍창기,정윤철 한국물리학회 2020 THE JOURNAL OF THE KOREAN PHYSICAL SOCIETY Vol.76 No.2
The conductance through a QPC {is calculated} when a hill like impurity potential, due to the nonuniform distribution of ionized impurities in the modulation doping layer, is added to smooth varying saddle like QPC potential. It was found that a weakly localized state can form between the QPC center and the impurity. Anomalous plateau appeared around 0.7 × 2e^{2}/h when the conductance via the localized state is added to the QPC conductance. The results imply that the resonance due to quasi-bound state formed by the impurity potential in the QPC is one of the possible origins of anomalous conductance plateau.
홍창기(Changgi Hong),권기현(Gihwon Kwon) 한국정보기술학회 2022 Proceedings of KIIT Conference Vol.2022 No.12
강화학습 알고리즘은 특정 환경에 대해 에이전트를 훈련시켜 하나의 에피소드에서 누적되는 보상의 합을 최대화하는 최적의 정책을 찾는다. 본 논문에서는 일반적인 강화학습의 목적에 반하는, 누적되는 보상의 합을 최소화하는 이른바 반항적 정책을 Q-Learning, SARSA 두 알고리즘을 통해 실패 도달 성능 차이를 보인다. 실험은 Open AI Gym의 FrozenLake 환경에서 기존의 학습방식은 유지한 채, 목표 상태가 아닌 실패 상태를 추적하도록 평가 과정을 변형하였다. 학습 과정과 평가 과정에서 보상, 총 보상 변화 추이와 실패 상태 도달경로 길이를 비교하였고, 그 결과 SARSA 알고리즘이 성능상 우세함을 실험을 통해 확인한다. 또한 두 알고리즘이 산출한 Q-테이블 내 데이터를 분석하여 해당 결과를 해석한다. Reinforcement learning algorithms find optimal policies that train agents at specific environment to maximize the sum of the accumulated rewards in one episode. In this paper, the so-called rebellious policy that minimizes the sum of accumulated rewards, contrary to the general purpose of reinforcement learning, shows the difference in failure reachabillity’s performance through two algorithms: Q-Learning and SARSA. The experiment transformed the evaluation process to track the failure state, not the goal state, while maintaining the existing learning method in the FrozenLake environment of Open AI Gym. In the learning and evaluation process, reward and total reward per episode changes, and the length of the path to fail state were compared, and as a result, it is confirmed through experiments that the SARSA algorithm is superior in performance. In addition, the results are interpreted by analyzing the data in the Q-table calculated by the two algorithms.
홍창기,변재용,여승근,차창일,박문서 대한이비인후과학회 2007 대한이비인후과학회지 두경부외과학 Vol.50 No.2
Background and Objectives:Facial palsy in children is uncommon, but not rare. Most patients with facial palsy are idiopathic, as in Bels palsy. Nevertheles, other more serious causes should be carefully excluded before Bels palsy is diagnosed. We tried to characterize the causes, treatment methods, and the recovery rate of facial palsy in children. Subjcets and Method:Between patients ages ranged from 0 to 15. Results:The highest age groups were children in their school period. The causes of facial palsy in children were in the order of following frequences:Bells palsy (66.2%), infection (14.6%), trauma (13.4%), birth trauma (3.2%), leukemia (1.3%), facial burn (0.6%), iatrogenic (0.6%). Treatment methods according to causes were steroid, Bells palsy, 100% in birth trauma, 100% in acute otitis media, 83.3% in herpes zoster oticus, 33.3% in temporal bone fracture, and 50% in cerebral contusion. Conclusion:Facial palsy in children results from variable causes in the order of Bells palsy (66.2%), infection (14.6%), trauma (13.4%), birth trauma (3.2%), leukemia (1.3%), facial burn (0.6%), and iatrogenic (0.6%). There were variable treatment methods upuncture. Facial palsy caused by birth trauma and acute otitis media had satisfactory recovery rates in all cases, whereas Bels palsy patients showed 93.1% satisfactory recovery and trauma patients 33.3%. (Korean J Otolaryngol 207 ;50 :102-7)