• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.41 no.6
• /
• pp.25-31
• /
• 2004

In this paper, we introduce kinetic Monte Carlo (kMC) methods for simulating diffusion process in nano-scale device fabrication. At first, we review kMC theory and backgrounds and give a simple point defect diffusion process modeling in thermal annealing after ion (electron) implantation into Si crystalline substrate to help understand kinetic Monte Carlo methods. kMC is a kind of Monte Carlo but can simulate time evolution of diffusion process through Poisson probabilistic process. In kMC diffusion process, instead of. solving differential reaction-diffusion equations via conventional finite difference or element methods, it is based on a series of chemical reaction (between atoms and/or defects) or diffusion events according to event rates of all possible events. Every event has its own event rate and time evolution of semiconductor diffusion process is directly simulated. Those event rates can be derived either directly from molecular dynamics (MD) or first-principles (ab-initio) calculations, or from experimental data.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.9 no.6
• /
• pp.1341-1345
• /
• 2005

Sequential Monte Carlo techniques are a set of powerful and versatile simulation-based methods to perform optimal state estimation in nonlinear non-Gaussian state-space models. We can use Monte Carlo particle filters adaptively, i.e. so that they simultaneously estimate the parameters and the signal. However, Sequential Monte Carlo techniques require the use of special panicle filtering techniques which suffer from several drawbacks. We consider here an alternative approach combining particle filtering and Sequential Hybrid Monte Carlo. We give some examples of applications in fisheries(luespotted mud hopper data).

• The KIPS Transactions:PartB
• /
• v.8B no.3
• /
• pp.231-236
• /
• 2001

일반적인 다층 신경망에서 가중치의 갱신 알고리즘으로 사용하는 오류 역전과 방식은 가중치 갱신 결과를 고정된(fixed) 한 개의 값으로 결정한다. 이는 여러 갱신의 가능성을 오직 한 개의 값으로 고정하기 때문에 다양한 가능성들을 모두 수용하지 못하는 면이 있다. 하지만 모든 가능성을 확률적 분포로 표현하는 갱신 알고리즘을 도입하면 이런 문제는 해결된다. 이러한 알고리즘을 사용한 베이지안 신경망 모형(Bayesian Neural Networks Models)은 주어진 입력값(Input)에 대해 블랙 박스(Black-Box)와같은 신경망 구조의 각 층(Layer)을 거친 출력값(Out put)을 계산한다. 이 때 주어진 입력 데이터에 대한 결과의 예측값은 사후분포(posterior distribution)의 기댓값(mean)에 의해 계산할 수 있다. 주어진 사전분포(prior distribution)와 학습데이터에 의한 우도함수(likelihood functions)에 의해 계산한 사후확률의 함수는 매우 복잡한 구조를 가짐으로 기댓값의 적분계산에 대한 어려움이 발생한다. 따라서 수치해석적인 방법보다는 확률적 추정에 의한 근사 방법인 몬테 칼로 시뮬레이션을 이용할 수 있다. 이러한 방법으로서 Hybrid Monte Carlo 알고리즘은 좋은 결과를 제공하여준다(Neal 1996). 본 논문에서는 Hybrid Monte Carlo 알고리즘을 적용한 신경망이 기존의 CHAID, CART 그리고 QUEST와 같은 여러 가지 분류 알고리즘에 비해서 우수한 결과를 제공하는 것을 나타내고 있다.

• The Korean Journal of Applied Statistics
• /
• v.13 no.1
• /
• pp.189-196
• /
• 2000

Conjugate prior density families were motivated by considerations of tractability in implementing the Bayesian paradigm. But we consider problem that the conjugate prior p($\Theta$) cannot be used in restriction of the parameter $\Theta$. This article considers the nonconjugate prior problem of hierarchical Poisson model. We demonstrate the use of latent variables for sampling non-standard densities which arise in the context of the Bayesian analysis of non-conjugate by using a Gibbs sampler.

• KIISE Transactions on Computing Practices
• /
• v.21 no.1
• /
• pp.94-99
• /
• 2015

With the recent machine learning paradigm of using nonparametric Bayesian statistics and statistical inference based on random sampling, the Dirichlet distribution finds many uses in a variety of graphical models. It is a multivariate generalization of the gamma distribution and is defined on a continuous (K-1)-simplex. This paper presents a sampling method for a Dirichlet distribution for the problem of dividing an integer X into a sequence of K integers which sum to X. The target samples in our problem are all positive integer vectors when multiplied by a given X. They must be sampled from the correspondingly gridded simplex. In this paper we develop a Markov Chain Monte Carlo (MCMC) proposal distribution for the neighborhood grid points on the simplex and then present the complete algorithm based on the Metropolis-Hastings algorithm. The proposed algorithm can be used for the Markov model, HMM, and Semi-Markov model for accurate state-duration modeling. It can also be used for the Gamma-Dirichlet HMM to model q the global-local duration distributions.

• Proceedings of the IEEK Conference
• /
• 2000.06b
• /
• pp.326-329
• /
• 2000

In this paper, a complete methodology of incorporating the displacement current for the calculation of a single electron inverter characteristics has been devised. It has been implemented for the calculation of the low frequency noise spectrum in a single electron inverter in the framework of Monte-Carlo method. Our new methodology opens up a systematic way of analyzing transient behaviors of single electron circuits.

• Journal of Internet Computing and Services
• /
• v.21 no.5
• /
• pp.1-7
• /
• 2020

Deep neural networks(DNN), which are used as approximation functions in reinforcement learning (RN), theoretically can be attributed to realistic results. In empirical benchmark works, time difference learning (TD) shows better results than Monte-Carlo learning (MC). However, among some previous works show that MC is better than TD when the reward is very rare or delayed. Also, another recent research shows when the information observed by the agent from the environment is partial on complex control works, it indicates that the MC prediction is superior to the TD-based methods. Most of these environments can be regarded as 5-step Q-learning or 20-step Q-learning, where the experiment continues without long roll-outs for alleviating reduce performance degradation. In other words, for networks with a noise, a representative network that is regardless of the controlled roll-outs, it is better to learn MC, which is robust to noisy rewards than TD, or almost identical to MC. These studies provide a break with that TD is better than MC. These recent research results show that the way combining MC and TD is better than the theoretical one. Therefore, in this study, based on the results shown in previous studies, we attempt to exploit a random balance with a mixture of TD and MC in RL without any complicated formulas by rewards used in those studies do. Compared to the DQN using the MC and TD random mixture and the well-known DQN using only the TD-based learning, we demonstrate that a well-performed TD learning are also granted special favor of the mixture of TD and MC through an experiments in OpenAI Gym.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.9 no.4
• /
• pp.193-203
• /
• 2001

The Monte Carlo method was used to evaluate the reliability of crash discrimination algorithms. Through the Fast Fourier Transformation, crash pulses obtained during frontal crash tests of a mini van and a sports utility vehicle were transformed to signals in the frequency domain, and the signals were divided into basic signals and changeable signals. The changeable signals were modified through random generation, and they were combined with the basic signals. Then, the combined signals were transferred back to the time domain. In this way numerous crash pulses could be generated. For the generated pulses, crash discrimination algorithms were evaluated by examining whether they did not result in air bag deployment for the pulses requiring no air bag deployment and whether they resulted in time-to-fires faster than required time-to-fires for the pulses requiring air bag deployment. The crash discrimination algorithm in which the absolute value of the deceleration change multiplied by the velocity change or the summation of the absolute value of the deceleration change was used as a metric was Proven to be highly reliable.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.2
• /
• pp.40.2-40.2
• /
• 2019

허블 텐션이란 허블우주망원경으로 관측한 허블상수 값과 플랑크 위성으로 측정한 허블상수 값이 일치하지 않는 문제를 일컬으며 현재 우주론에서 주목 받는 이슈 중 하나이다. 밀도가 작은 지역에선 약한 중력으로 공간의 팽창이 빠르고, 반대로 밀도가 큰 지역에서는 팽창이 느리다. 만약, 우리 근처에서 상대적으로 낮은 밀도 때문에 팽창 속도의 차이가 생긴다면 허블 텐션의 원인을 쉽게 설명할 수 있다. 이 문제를 구체적으로 다루기 위해, 우리는 우주 상수를 고려한 아인슈타인 중력의 구형 우주론 풀이인 Lambda-Lemaître-Tolman (ΛLT) 모형을 사용하였다. 우리로부터 먼 현상은 기존의 ΛCDM(Λ cold dark matter) 모형으로, 가까운 현상은 국소적인 LT 모형으로 기술함으로써 허블 텐션 문제를 해결하고자 하였다. 또한, 마코프 체인 몬테 칼로 (MCMC) 방법을 적용하여 천문 관측 결과를 잘 맞추는 ΛLT 모형의 변수들을 탐색하였다.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.47 no.8
• /
• pp.574-581
• /
• 2019

This paper investigates the effects of thrust misalignments and offsets of the lateral center of gravity of a space launch vehicle on its guidance performance. Sensitivity analysis and Monte Carlo simulations are applied to analyze their effects by computing changes in orbit injection errors when including the error sources. To compensate their effects, the attitude controller including an integrator additionally and the Steering Misalignment Correction (SMC) routine of the Saturn V are considered, and then Monte Carlo simulations are performed to evaluate their performances.