본문 바로가기 주메뉴 바로가기
22대 대한민국 국회 국회정보길라잡이 국회의원검색
회원가입 로그인
HOME

정보검색

소장정보 검색

국회도서관 홈으로 정보검색 소장정보 검색
결과 내 검색 동의어 포함
결과 내 검색 동의어 포함

-

목차보기

Title Page

Abstract

Contents

1. Identifying Soft Particles by Using Machine Learning Method in Glassy Liquid System 9

1.1. Introduction 9

1.1.1. Support Vector Machine 11

1.2. Simulation method for glassy liquid system 20

1.3. Identifying particle rearrangements through Phop[이미지참조] 21

1.4. Training method and result of the support vector machine 26

1.5. Comparing SVM with Neural Network 32

1.6. Conclusion 33

2. Local Structural Signature of Cooperative Dynamics in Glassy Liquid System 34

2.1. Introduction 34

2.2. Results 36

2.2.1. Origin of the heterogeneity 40

2.2.2. Decomposition of instanton time 42

2.2.3. Rationalize the local excitations: Change in structure as response to the local motion 49

2.2.4. Relation between softness and inherent structure of particles 54

2.3. Conclusion 61

3. Image Charge Induced Morphology of Water Droplet 62

3.1. Introduction 62

3.2. Onsager's principle 64

3.3. Profile evolution of the droplet 67

3.3.1. Dissipation function 69

3.3.2. Free energy variation 70

3.4. Results 73

3.5. Conclusion 75

References 76

List of Figures

Figure 1-1. The description of the support vectors and the decision boundaries are shown. The... 12

Figure 1-2. The slack variable ξ is described in the figure. In the C-SVM, data points can exist on... 15

Figure 1-3. The idea of the mapping function of the kernel SVM is shown in the figure. The figure... 18

Figure 1-4. The dependence of the average duration of rearrangement on maximum Phop. The...[이미지참조] 22

Figure 1-5. The dependence of the average displacement size of rearrangement on maximum Phop....[이미지참조] 23

Figure 1-6. The distribution of the displacement size, according to the different maximum Phop...[이미지참조] 24

Figure 1-7. The distribution of the duration of rearrangement, according to the different maximum... 25

Figure 1-8. The distribution of softness with different Phop threshold for the soft label. The shape...[이미지참조] 27

Figure 1-9. The configuration of high softness with Phop,training＝0.2.[이미지참조] 28

Figure 1-10. The configuration of high softness with Phop,training＝0.6.[이미지참조] 29

Figure 1-11. The configuration of low softness with Phop,training＝0.2.[이미지참조] 30

Figure 1-12. The configuration of low softness with Php,training＝0.6.[이미지참조] 31

Figure 1-13. The dimesionality of each layers in the neural network is described in the figure. 32

Figure 2-1. Classification of particles by using a support vector machine: soft(red) and hard(blue) particles. 36

Figure 2-2. The distribution of particles with high softness. 37

Figure 2-3. The distribution of particles with low softness. 38

Figure 2-4. When a particle escapes from its local cage, Phop is large and │ri(t)-ri(0)│ change...[이미지참조] 39

Figure 2-5. (a) Four particles exhibit high mobility at the same time, (b) Their displacements are... 41

Figure 2-6. Examples of elemental processes of primary particles: a single primary particle. 43

Figure 2-7. Examples of elemental processes of primary particles: two primary particles moving along. 44

Figure 2-8. Examples of elemental processes of primary particles: two primary particles moving... 45

Figure 2-9. Examples of elemental processes of primary particles: three primary particles moving along. 46

Figure 2-10. Examples of elemental processes of primary particles: three primary particles display-... 47

Figure 2-11. Examples of elemental processes of primary particles: several primary particles... 48

Figure 2-12. Inherent structures of final state of MD simulation and fictitious process are plotted... 51

Figure 2-13. The average displacements of particles are plotted as functions of the distance from... 52

Figure 2-14. The number of particles as a function of average displacement is displayed as solid... 53

Figure 2-15. The trajectories of rearrangements with a small number of primary particles are... 55

Figure 2-16. Comparison with inherent structure, initial and final position of rearrangements... 56

Figure 2-17. The configuration of softness around primary particles and corresponding inherent... 57

Figure 2-18. The trajectories of rearrangements with a large number of primary particles are... 58

Figure 2-19. Comparison with inherent structure, initial and final position of rearrangements... 59

Figure 2-20. The configuration of softness around primary particles and corresponding inherent... 60

Figure 3-1. (a) The description of the droplet system, when the distance between upper plate and... 68

Figure 3-2. The evolution of height profile is shown. The unit of time is given as second. The length... 73

Figure 3-3. The shape evolution of the top side of the droplet. The unit of time is given in second.... 74

초록보기

The rapid increase of the computing power and the development of the efficient machine learning techniques in recent days allow us to apply the machine learning to various field of researches. The machine learning techniques become effective solutions to the fields which are not well described by one integrated theory. The physics for the super cooled glassy liquid system is one such field. The prediction of local dynamics based on the local structural information is longstanding problem in glass physics. The latest researches reveal that the local dynamics is actually predictable by utilizing the support vector machine (SVM) and the local structural information. In this thesis, I reproduce the result of the former researches which utilize the SVM and compare it with the result of the neural network. In addition, I suggest the origin of the dynamic heterogeneity in glassy liquid system. Furthermore, I demonstrate the composition of the local soft regions in glassy liquid systems.

추천서가 (다양한 추천 자료를 만나보세요)

권호기사보기

권호기사 목록 테이블로 기사명, 저자명, 페이지, 원문, 기사목차 순으로 되어있습니다.
기사명 저자명 페이지 원문 기사목차