• Korean Journal of Logic
• /
• v.8 no.1
• /
• pp.23-46
• /
• 2005

The purpose of the paper Is to discuss and estimate early Popper's theory of probability, which is presented in his book, The Logic of of Scientific Discovery. For this, Von Mises' frequency theory shall be discussed in detail, which is regarded as the most systematic and sophisticated frequency theory among others. Von Mises developed his theory to response to various critical questions such as how finite and empirical collectives can be represented in terms of infinite and mathematical collectives, and how the axiom of randomness can be mathematically formulated. But his theory still has another difficulty, which is concerned with the inconsistency between the axiom of convergence and the axiom of randomness. Defending the objective theory of probability, Popper tries to present his own frequency theory, solving the difficulty. He suggests that the axiom of convergence be given up and that the axiom of randomness be modified to solve Von Mises' problem. That is, Popper introduces the notion of ordinal selection and neighborhood selection to modify the axiom of randomness. He then shows that Bernoulli's theorem is derived from the modified axiom. Consequently, it can be said that Popper solves the problem of inconsistency which is regarded as crucial to Von Mises' theory. However, Popper's suggestion has not drawn much attention. I think it is because his theory seems anti-intuitive in the sense that it gives up the axiom of convergence which is the basis of the frequency theory So for more persuasive frequency theory, it is necessary to formulate the axiom of randomness to be consistent with the axiom of convergence.

• The Bulletin of The Korean Astronomical Society
• /
• v.38 no.1
• /
• pp.75.1-75.1
• /
• 2013

자체의 중력 효과를 고려하는 구대칭 완전 유체 전산모사 연구를 위해 일반상대론적 유체역학 코드를 이 분야 연구자들을 위한 공개용으로 개발하였다. 이 코드는 3+1 ADM(Arnowitt-Deser-Misner) 공식과 등방 공간 좌표를 사용하였다. 시공간 기하를 구하기 위해 극한값 썰기 (maximal slicing) 조건과 함께 세 개의 제한 방정식을 풀었고, 시공간을 채우는 물질인 유체는 근사 리만 해법을 사용한 HRSC (high resolution shock capturing) 기법으로 오일러 관찰자 시점에서 풀었다. 이 코드의 수렴성과 정확성을 검증하기 위해 상대론적인 구대칭 충격파 비교 분석, 블랙홀로 빨려 들어가는 상대론적 구대칭 강착, TOV(Tolman-Oppenheimer-Volkoff) 별 및 OS (Oppenheimer-Snyder) 붕괴 코드 테스트를 수행하였다. 특히, 이 코드의 동적 진화 테스트인 OS 붕괴의 경우 해석적인 해와 결과를 비교하기 위하여 좌표변환을 수치 계산으로 수행하였다. 아인슈타인의 일반상대성 이론을 넘어서는 변형된 중력이론 중 하나로 최근 제시된 EiBI(Eddington-inspired Born-Infeld) 이론에서 TOV 별의 해가 일반상대성 이론과 어떠한 차이를 보이는지 살펴 보았고, 그 이론에서도 물질이 붕괴하여 블랙홀을 만드는 경우 특이점이 형성되는지 고찰해 보았다.

• Journal of Energy Engineering
• /
• v.8 no.1
• /
• pp.67-75
• /
• 1999

여러 종류의 내화재로 구성된 석탄가스화기에서의 온도분포 및 열손실량을 구하기 위한 전산해석을 수행하였다. 석탄가스화기 내화재 설계를 위한 적절한 방법론을 제안하기 위하여 1차원 이론적 해석, 2차원 전도열전달 해석 및 3차원 대류-전도 복합열전달 해석 등 세가지 방법론으로 해석을 각각 수행하였다. 해석 결과들은 석탄가스화기 실험 결과와 정상적 정량적으로 잘 일치하는 것으로 나타났다. 결과의 정확성, 수치해석 상의 수렴성 및 계산시간 등을 종합적으로 고려해 볼 때, 전산해석에 핵심 경계조건인 가스화기 내벽의 온도를 적절히 설정할 수 있는 경우에는 2차원 전도열전달 해석이 공학적 설계에 적용하기 알맞은 방법론으로 판단되었다. 전산해석 결과에 의하면, 현재 실험이 진행중인 하루 3톤 처리 용량급의 석탄가스화기에서의 총 열손실량은 설계치 운전 기준으로 약 1% 정도인 것으로 판별되었다.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2014.05a
• /
• pp.682-685
• /
• 2014

In this paper, an improved designed of the circularly polarization antenna for sensor node. Stochastic optimization algorithms of Particle Swarm Optimization (PSO) and Adaptive Particle Swam Optimization(APSO) are studied and compared. To verify that the APSO is working better than the standard PSO, the design of a circularly polarization antenna is shows the optimized result with 27 iterations in the APSO and 41 iterations in th PSO.

• Proceedings of the KIEE Conference
• /
• 1999.07g
• /
• pp.2833-2835
• /
• 1999

이 논문에서는 임의의 시스템 동정에 사용되는 적응필터의 계수를 최적화시키는 방법으로 광범위하게 사용되어지고 있는 기존의 적응 알고리즘인 Least Mean Square(LMS)방법과 최근들어 다양한 최적화 문제에 응용되고 있는 유전자 알고리즘(GA)을 합성한 하이브리드 형태의 적응 알고리즘을 사용한다. 이 알고리즘은 TIR 필터를 설계하는데 있어, 경사하강법의 개념을 사용함으로써 야기되는 지역 수렴문제의 단점을 보완하기 위해, 미분과 같은 결정론적인 규칙없이 단지 확률적인 연산자만으로 진행하는 유전자 알고리즘을 이용한다. 그리고 유전자 알고리즘에 있어서 확률적인 연산을 사용함으로써 발생하는 많은 계산량과 느린 수렴속도 문제를 LMS의 경사하강법을 이용하여 보완한다. 이처럼 유전자 알고리즘이 지닌 장점과 LMS 알고리즘이 갖는 장점을 이용하여 각 알고리즘이 지니는 단점을 서로 보완함으로써 알고리즘의 성능을 향상시키고 이 향상된 알고리즘을 이용하여 최적 필터계수를 찾는다 이렇게 얻은 필터계수값을 이용하여 적응 필터의 성능을 확인 평가한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.10a
• /
• pp.217-219
• /
• 2004

본 연구에서는 Feedforward Neural Network에 적용될 수 있는 개선된 학습 알고리즘을 개발하고자 한다. 제시된 알고리즘을 이용하여 학습을 할 때 학습 초기는 가장 단순한 경우로써 한 개의 학습 패턴과 은닉 층으로부터 시작한다. 신경망 학습 중에 지역 최소값에 수렴되면 weights scaling 기법을 이용하여 지역 최소값을 벗어나도록 한다. 지역 최소값의 탈출이 용이하지 않으면 은닉노드를 점차적으로 추가한다. 이러한 단계에서 새롭게 추가된 노드에 대한 초기값 선택은 선형계획법을 이용한 최적 처리절차론 이용한다. 최적 처리절차의 결과로써 은닉 층의 노드가 추가된 후의 네트워크는 학습회수를 증가시키지 않아도 학습 허용오차를 만족시킬 수 있다 본 연구에서 적용한 개선된 알고리즘을 이용하면 신경망 학습시의 수렴 정도를 높여주고 최소한의 단순 구조를 갖는 신경망으로 추정할 수 있게 하며, 이 결과를 모의실험을 통하여 살펴보고 기존의 연구 결과와 비교한다.

• Journal of the Korean Society for Precision Engineering
• /
• v.14 no.2
• /
• pp.152-159
• /
• 1997

The asymptotic convergence of a coupled dynamic system, which is motivated from infinite dimensional adaptive systems, is investigated. The convergence analysis is formulated in abstract Banch spaces and is shown to applicable to a broad class of infinite dimensional systems including adaptive identification and adaptive control. Particularly it is shown that if a uniquely existing solution is p-th power integrable, then the solution converges to zero asymptotically. The persistence of excitation(PE) of a signal which arises in an infinite dimensional adaptive system is investigated. The PE property is not completely known yet for infinite dimensional adaptive systems, however it should be investigated in relation to spatial variable, boundary conditions as well as time variable.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.8 no.5
• /
• pp.98-106
• /
• 1998

This paper proposes an efficient learning algorithm for improving the training performance of the neural network. The proposed method improves the training performance by applying the backpropagation algorithm of a global optimization method which is a hybrid of a stochastic approximation and a conjugate gradient method. The approximate initial point for f a ~gtl obal optimization is estimated first by applying the stochastic approximation, and then the conjugate gradient method, which is the fast gradient descent method, is applied for a high speed optimization. The proposed method has been applied to the parity checking and the pattern classification, and the simulation results show that the performance of the proposed method is superior to those of the conventional backpropagation and the backpropagation algorithm which is a hyhrid of the stochastic approximation and steepest descent method.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.3
• /
• pp.197-208
• /
• 2023

The purpose of this study is to establish definitions on Metaverse and Metaverse literacy and to develop a framework for Metaverse literacy utilizing a Delphi study and in-depth interview methods with experts on Metaverse and digital, and future education. 9 experts participated in the Delphi study and 4 took part in the in-depth interview, and Delphi study data was collected via emails from January 12th to 17th and in-depth study data was gathered via zoom from March 1st to 19th in 2022. To establish the definition of Metaverse, 6 existing definitions were presented and ranked with calculated means of each definition, and additional comments were collected. To build the definition of Metaverse literacy, 1 existing definition was shown and calculated with the degree of agreement, and additional feedback was gathered. To develop a framework for Metaverse literacy, 12 sub-concepts were presented and ranked with calculated means of each concept. After applying feedback of the Delphi study, the authors re-established the definitions of Metaverse and Metaverse literacy, and re-develop the framework for Metaverse literacy. In the in-depth interview, feedback on these definitions and the framework was collected. After these processes, the definitions of Metaverse and Metaverse literacy were finalized, and the framework for Metaverse literacy was developed. Based on the results, implications and limitations of the study, and possible future study are discussed.

• Journal of Science Education
• /
• v.32 no.2
• /
• pp.121-145
• /
• 2008

The purposes of this study were to identify the misconceptions that students have on the magma and plate tectonics and to present the implications in developing textbooks as well as related curriculum of high school textbooks. Data were collected through questionnaire, consisting of some questions, short essays, and descriptive drawings, developed by the research team. A total of 140 high school students(9th graders) responded to those questionnaires and were interviewed for further information. It was reported that participants displayed various misconceptions related to magma and plate tectonics. The identified misconceptions are as follows: For the definition of magma, the 31% of participants misunderstood magma as lava. In respect to the generative mechanism of magma at subduction zone, over 90% of students responded that it is generated by frictional heat. The source of misconceptions were identified as a result from textbooks and related reference-books. For the concept of plates, 87% of students conceived 'crust or a lower part of the plates' as 'plates'. Most participants hold the right concept of oceanic ridge, whereas, 66% of them considered 'rift valley' as either 'divergence of continental plates' or 'converging boundary'. 63% of them defined 'collision boundary of continental plate' as either 'subduction zone' or 'diverging boundary'. For the definitions of the trench and Benioff zone, 86% of students responded them as the place of subduction or differing density between two converging plates. The students' misconceptions were resulted from the errors and insufficient explanation, inappropriate figures, and data presented in textbooks, reference-books, lecture, and web sites. The results of this study are implied to contribute the improvement of students' misconceptions.