• 인터넷정보학회논문지
• /
• 제10권3호
• /
• pp.9-15
• /
• 2009

보드게임은 많은 수의 말들과 상태공간을 갖고 있다. 그래서 학습이 많은 시간동안 학습을 하여야 한다. 또한 상대방과의 대결이 1 대 1 로 이루어지지 않고, 여러 말 대 여러 말로 이루어지므로 전략적인 사고가 필요하다. 그러므로 최적의 학습을 적용하여야 한다. 본 논문에서는 강화학습 알고리즘을 이용하였다. 보상 값을 받아 보드게임 말이 학습하게 하여 지능적으로 움직이게 하였다. 학습 도중에 동일한 최선 값이 있을 때, 줄고누 문제 영역 지식을 활용한 휴리스틱을 사용해 학습의 속도 향상을 시도하였다. 단순 구현된 말과 개선 구현된 말을 비교하기 위해 보드게임을 제작하였다. 그래서 일방공격형 말과 승부를 하게 하였다. 실험결과 개선 구현된 말의 성능이 학습속도 측면에서 월등히 향상됨을 알 수 있었다.

• 인터넷정보학회논문지
• /
• 제12권3호
• /
• pp.77-88
• /
• 2011

가용 주파수 고갈 문제를 해결하기 위하여 제안된 무선인지기술은 특정 주파수 대역에 대해 사용면허를 가진 주사용자가 사용하지 않는 유휴채널에 접근하여 통신을 수행함으로써 주파수 효율을 향상시키는 차세대 통신기술이다. 주사용자의 유휴채널을 사용하기 위해서는 해당 채널을 현재 주사용자가 점유하고 있는지를 정확히 판단하여야 한다. 분산형 무선인지 네트워크에서 독립적으로 채널을 센싱하는 무선인지 기기의 경우 센싱의 결과가 노이즈, 쉐도윙, 페이딩과 같은 채널 환경에 영향을 많이 받으며 심지어 주사용자가 요구하는 간섭률을 보장하지 못하는 결과를 초래한다. 따라서 본 논문에서는 주사용자가 요구하는 최소 간섭량을 보장하는 동시에 기회주의적으로 채널에 접근하여 인지시스템의 처리율(처리율)을 향상시키는 Q-learning 기반의 채널접근기법을 제안한다. 제안하는 기법은 사전 학습 단계에서 주사용자의 채널사용 패턴을 Q-learning으로 학습하고 이를 Q-learning 기반 채널접근 단계에서 실제로 적용함으로써 스펙트럼 센싱 성능을 향상시킨다. 모의실험을 통해 AWGN 및 레일레이 페이딩 무선 환경에서 주사용자에 대한 간섭량 및 처리율 성능이 기존의 에너지 검출 방법에 비해 더 우수함을 확인하였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권4호
• /
• pp.2038-2057
• /
• 2017

Gaussian learning is a new technology in the computational intelligence area. However, this technology weakens the learning ability of a particle swarm and achieves a lack of diversity. Thus, this paper proposes a vector Gaussian learning strategy and presents an effective approach, named particle swarm optimization based on vector Gaussian learning. The experiments show that the algorithm is more close to the optimal solution and the better search efficiency after we use vector Gaussian learning strategy. The strategy adopts vector Gaussian learning to generate the Gaussian solution of a swarm's optimal location, increases the learning ability of the swarm's optimal location, and maintains the diversity of the swarm. The method divides the states into normal and premature states by analyzing the state threshold of the swarm. If the swarm is in the premature category, the algorithm adopts an inertia weight strategy that decreases linearly in addition to vector Gaussian learning; otherwise, it uses a fixed inertia weight strategy. Experiments are conducted on eight well-known benchmark functions to verify the performance of the new approach. The results demonstrate promising performance of the new method in terms of convergence velocity and precision, with an improved ability to escape from a local optimum.

• 한국융합학회논문지
• /
• 제10권5호
• /
• pp.9-16
• /
• 2019

본 논문에서는 딥러닝 기술이 인터넷과 연결된 다양한 비즈니스 분야에 새로운 형태의 비즈니스 업무에 활용할 수 있도록 보안에 관한 문제점을 분석하고자 한다. 우선 딥러닝이 비즈니스 영역에 보안 업무를 충분히 수행하기 위해서는 많은 데이터를 가지고 반복적인 학습을 필요하게 된다. 본 논문에서 딥러닝이 안정적인 비즈니스 보안 업무를 완벽하게 수행할 수 있는 학습적 능력을 얻기 위해서는 비정상 IP패킷에 대한 탐지 능력과 정상적인 소프트웨어와 악성코드를 탑재하여 감염 의도를 가지고 접근하는 공격을 탐지해낼 수 있는 인지 능력을 갖추고 있는지를 분석하였다. 이에 본 논문에서는 인공지능의 딥러닝 기술이 시스템에 접근하여 문제의 비즈니스 모델을 안정적으로 수행할 수 있게 하기 위해서는 시스템내의 비정상 데이터를 추출해 내고 시스템 데이터 침해를 구분해 낼 수 있는 수학적 역할의 문제점을 보완하기 위해 새로운 IP에 대한 세션 및 로그 분석을 수행할 수 있도록 보안 엔진이 탑재된 딥러닝 기술을 개발하여 비즈니스 모델에 적용시켜서 취약점을 제거하여 비즈니스 업무 능력을 향상시키도록 문제적 방안을 비교 분석하였다.

• 대한간호학회지
• /
• 제3권2호
• /
• pp.91-100
• /
• 1973

I. Objectives of the study This study attempted In examine, from a pedagogical point of view, how socially-handicapped children differ from normal ones in their family backgrounds, personality-traits. adaptability to school life, and their peer relationship. This study was made under the following hypothesis; 1. The family background of socially-handicapped children is less desirable than that of normal ones. 2. Personality traits of socially-handicapped children are less desirable than those of normal ones. 3. Socially handicapped children tend to be less adaptable to school life than normal ones. 4. Peer-relationship of socially-handicapped children is less desirable than of normal ones . II. Contents of study The thesis consists of the following five main parts; 1. Introduction 2. Related Studies 3. Procedure 4. Results 5. Summary and Conclusions III. Instruments The following instruments were used for this study; 1. Family background record prepared by the school. 2. Questionnaire prepared by the writer. 3. General personality test (written by Kim Ki-Suk and published by Korean Testing Center). 4. School activity record. 5. Sociocratic test. IV. The following conclusions were derived from the study 1, As compared with normal children, socially-handicapped ones have, in many cases, larger number of siblings, their families belong to lower economic bracket: their parents were indifferent to their children, and their mothers were less educated. 2. As compared with normal children, socially-handicapped ones are emotionally unstable, although they are the same as normal children in their sense of dominancy, responsibility, sociality, and confidence. 3. As compared with normal children, socially-handicapped ones are tardy in their learning, although they are almost the same as normal ones in their intelligence development. 4. As compared with normal children, socially-handicapped ones make friends more easily with pupils of other classes and schools rather than those of their own class or school.

• Investigative Magnetic Resonance Imaging
• /
• 제9권2호
• /
• pp.117-133
• /
• 2005

자기공명영상에서 병리구조물을 깨닫기 위해서는 먼저 자기공명영상에서 정상 해부구조물을 깨달아야 한다. 자기공명영상에서 해부구조물을 익히기 위해서는 다음과 같은 학습 자료가 필요하다. 첫째, 온몸의 자기공명영상, 둘째, 수평, 이마 그리고 마루 자기공명영상, 셋째, 자기공명영상에 들어맞는 구역화영상, 넷째, 자기공명영상에 있는 해부구조물의 3차원영상, 넷째, 자기공명영상, 구역화영상 그리고 3차원영상을 볼 수 있는 소프트웨어가 필요하다. 그러나 지금까지 이러한 학습 자료를 구하기 힘들었다. 따라서 이 연구에서는 의과대학 학생과 의사가 자기공명영상에서 정상 해부구조물을 익히는 데 도움을 주는 학습 자료를 다음처럼 만들었다. 표준 체형을 가진 건강한 한국인 남성을 골랐다. 온몸의 자기공명영상 613장을 찍고(slice thickness 3 mm, interslice gap 0 mm, field of view 480mm${\times}$480mm, resolution 512${\times}$512, T1 weighted) 개인용 컴퓨터에 옮겼다. 자기공명영상에 있는 60개의 해부구조물을 구역화해서 구역화영상을 만들었다. 이마, 마루 자기공명영상과 이마, 마루 구역화영상을 만들었다. 구역화영상을 바탕으로 47개 해부구조물의 3차원영상을 수동 표면재구성 방법으로 만들었다. 자기공명영상, 구역화영상, 3차원영상을 볼 수 있는 소프트웨어를 만들었다. 이 연구에서 만든 온몸의 수평, 이마, 마루 자기공명영상, 자기공명영상에 들어맞는 구역화영상, 3차원영상, 소프트웨어와 같은 학습 자료는 의과대학 학생과 의사가 자기공명영상에서 정상 해부구조물을 익히는 데 도움을 줄 것이다. 이 학습 자료는 인터넷이나 CD를 통해서 널리 퍼뜨릴 것이다.

• Animal cells and systems
• /
• 제7권4호
• /
• pp.303-307
• /
• 2003

The aim of this study is to investigate whether alcohol alters learning and memory processes pertaining to emotional and spatial factors using the active avoidance and T-maze task in zebrafish. In the active avoidance task, zebrafish were trained to escape from one compartment to another to avoid electric shocks (unconditioned stimulus) following a conditioned light signal. Acquisition of active avoidance task appeared to be normal in zebrafish that were treated with 1% alcohol for 30 min for 17 days until the end of the behavioral test, and retention ability of learned behavior, tested 2 days later, was the same as control group. In the T-maze task, the time to find a reservoir was compared. While the latency was similar during the 1 st training session between control and alcohol-treated zebrafish, it was significantly longer in alcohol-treated zebrafish during retention test 24 h later. Furthermore, when alcohol was treated 30 min after 2nd session without prior treatment, zebrafish demonstrated similar retention ability compared to control. These results suggest that chronic alcohol treatment alters spatial learning of zebrafish, but not emotional learning.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권9호
• /
• pp.3762-3781
• /
• 2020

Infrared pedestrian target recognition is a vital problem of significant interest in computer vision. In this work, a novel infrared pedestrian target recognition method that uses heterogeneous features with multi-kernel transfer learning is proposed. Firstly, to exploit the characteristics of infrared pedestrian targets fully, a novel multi-scale monogenic filtering-based completed local binary pattern descriptor, referred to as MSMF-CLBP, is designed to extract the texture information, and then an improved histogram of oriented gradient-fisher vector descriptor, referred to as HOG-FV, is proposed to extract the shape information. Second, to enrich the semantic content of feature expression, these two heterogeneous features are integrated to get more complete representation for infrared pedestrian targets. Third, to overcome the defects, such as poor generalization, scarcity of tagged infrared samples, distributional and semantic deviations between the training and testing samples, of the state-of-the-art classifiers, an effective multi-kernel transfer learning classifier called MK-TrAdaBoost is designed. Experimental results show that the proposed method outperforms many state-of-the-art recognition approaches for infrared pedestrian targets.

• 대한간호학회지
• /
• 제38권1호
• /
• pp.152-160
• /
• 2008

Purpose: It has been attempted to support mother of premature infants by providing information of premature infant care using e-Learning because premature infants need continuous care from birth to after discharge. Method: The e-Learning Program for mother of premature was developed with Xpert, Namo web editor, Adobe Photoshop, and PowerPoint and applied for 4 weeks from 4 to 30 September 2006. Result: 1) We found that the contents of information which premature infants' need when being in the hospital and after discharge were the definition of a premature infant, orientation of NICU, care of premature infants, care of premature infants' common diseases, the connection of healthcare resources, exchange of information, and the management of rearing stress. 2) The program content consisted of cause of premature birth, comparison to full-term baby, physiology character, orientation of NICU, common health problems, follow up care, infection control, feeding, normal development physically and mentally, weaning method, and vaccination. Conclusion: Considering the results, this program for mother of premature is a useful means to provide premature-care information to mothers. This information can be readily accessible and can be varied and complex enough to be able to help mothers to the information and assistance they require.

• 시스템엔지니어링학술지
• /
• 제15권2호
• /
• pp.72-78
• /
• 2019

Pathology is the motor that drives healthcare to understand diseases. The way pathologists diagnose diseases, which involves manual observation of images under a microscope has been used for the last 150 years, it's time to change. This paper is specifically based on tumor detection using deep learning techniques. Pathologist examine the specimen slides from the specific portion of body (e-g liver, breast, prostate region) and then examine it under the microscope to identify the effected cells among all the normal cells. This process is time consuming and not sufficiently accurate. So, there is a need of a system that can detect tumor automatically in less time. Solution to this problem is computational pathology: an approach to examine tissue data obtained through whole slide imaging using modern image analysis algorithms and to analyze clinically relevant information from these data. Artificial Intelligence models like machine learning and deep learning are used at the molecular levels to generate diagnostic inferences and predictions; and presents this clinically actionable knowledge to pathologist through dynamic and integrated reports. Which enables physicians, laboratory personnel, and other health care system to make the best possible medical decisions. I will discuss the techniques for the automated tumor detection system within the new discipline of computational pathology, which will be useful for the future practice of pathology and, more broadly, medical practice in general.