• IEIE Transactions on Smart Processing and Computing
• /
• v.4 no.4
• /
• pp.202-208
• /
• 2015

In this study, we propose a new inference algorithm for a multiclass Gaussian process classification model using a variational EM framework and the Laplace approximation (LA) technique. This is performed in two steps, called expectation and maximization. First, in the expectation step (E-step), using Bayes' theorem and the LA technique, we derive the approximate posterior distribution of the latent function, indicating the possibility that each observation belongs to a certain class in the Gaussian process classification model. In the maximization step, we compute the maximum likelihood estimators for hyper-parameters of a covariance matrix necessary to define the prior distribution of the latent function by using the posterior distribution derived in the E-step. These steps iteratively repeat until a convergence condition is satisfied. Moreover, we conducted the experiments by using synthetic data and Iris data in order to verify the performance of the proposed algorithm. Experimental results reveal that the proposed algorithm shows good performance on these datasets.

• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.723-725
• /
• 2004

An iterative robust control design for PZT using Gaussian wavelet networks is proposed. A Gaussian wavelet network with accurate approximation capability is employed to approximate the nonlinear hysteresis dynamics of PZT systems by using an iterative control algorithm. Depending on the finite number of wavelet basis functions which results in unavoidable approximation errors, a robust control law is provided to guarantee the stability of the closed-loop nano positioning system. Finally, the effectiveness of the robust control approach is illustrated through comparative simulations on a PZT.

• The Korean Journal of Rheology
• /
• v.4 no.2
• /
• pp.127-137
• /
• 1992

묽은 고분자 용액의 유변학적 거동을 elastic dumbbell 모델을 사용하여 연구하면서 hydrodynamic interaction(H.I) 효과를 주로 살펴보았다. 먼저 consistent averaging방법을 사용하면서 Oseen tensor와 Rotne-Prager-Yama-kawa(R-P-Y) tensor를 H. I. tensor로 각 각 사용하여 그차이를 비교하였으며 oseen tensor를 사용하는 경우 tti-nger의 알고리듬과 Ahn과 Lee의 알고리듬을 비교하였다. 또 H.I. tensor를 처리하는 방법으로 consistent averaging 방법과 Gaussian approximation 방법의 차이에 대하여도 살펴보았다. Ahn과 Lee 의 알고리듬이 ttinger의 알고리듬보다 훨씬 빠른 계산시간을 보여주었으며 Gaussian approximation 방법을 사용하는 경우 consistent averaging 방법과 달리 second normal stress coefficient가 음의 값을 보이므로 더 합리적인 방법으로 생각된다.

• Management Science and Financial Engineering
• /
• v.19 no.2
• /
• pp.37-42
• /
• 2013

This paper suggests a numerical method for valuation of European and American options under the two L$\acute{e}$vy Processes, Normal Inverse Gaussian Model and the Variance Gamma model. The method is based on approximation of underlying asset price using a finite-state, time-homogeneous Markov chain. We examine the effectiveness of the proposed method with simulation results, which are compared with those from the existing numerical method, the lattice-based method.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.9 no.5
• /
• pp.1249-1254
• /
• 2008

The crucial part of graphical model is to compute the posterior distribution of parameters plus with the hidden variables given the observed data. In this paper, implementation of variational Bayes method for Gaussian mixture model and derivation of factorial variational approximation have been proposed. This result can be used for data analysis tasks like information retrieval or data visualization.

• ETRI Journal
• /
• v.33 no.6
• /
• pp.914-923
• /
• 2011

This paper concerns robust and reliable speaker model training for text-independent speaker verification. The baseline speaker modeling approach is the Gaussian mixture model (GMM). In text-independent speaker verification, the amount of speech data may be different for speakers. However, we still wish the modeling approach to perform equally well for all speakers. Besides, the modeling technique must be least vulnerable against unseen data. A traditional approach for GMM training is expectation maximization (EM) method, which is known for its overfitting problem and its weakness in handling insufficient training data. To tackle these problems, variational approximation is proposed. Variational approaches are known to be robust against overtraining and data insufficiency. We evaluated the proposed approach on two different databases, namely KING and TFarsdat. The experiments show that the proposed approach improves the performance on TFarsdat and KING databases by 0.56% and 4.81%, respectively. Also, the experiments show that the variationally optimized GMM is more robust against noise and the verification error rate in noisy environments for TFarsdat dataset decreases by 1.52%.

• Journal of the Korean Physical Society
• /
• v.73 no.11
• /
• pp.1612-1618
• /
• 2018

In this work, the ground-state properties of an interacting electron gas confined in a two-dimensional quantum dot system with the Gaussian potential ${\upsilon}(r)=V_0(1-{\exp}(-r^2/p))$, where $V_0$ and p are confinement parameters, are determined numerically by using the Thomas-Fermi approximation. The shape of the potential is modified by changing the $V_0$ and the p values, and the influence of the confining potential on the system's properties, such as the chemical energy, the density profile, the kinetic energy, the confining energy, etc., is analyzed for both the non-interacting and the interacting cases. The results are compared with those calculated for a harmonic potential, and excellent agreement is obtained in the limit of high p values for both the non-interacting and the interacting cases.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.10 no.3
• /
• pp.130-134
• /
• 1985

In two-dimensional discrete cosine transform(DCT) coding, the measurements of the distributions of the transform coefficients are important because a better approximation yields a smaller mean square distorition. This paper presents the results of distribution tests which indicate that the statistics of the AC coefficients are well approximated to a generalized Gaussian distribution whose shape parameter is 0.6. Furthermore, from a simulation of the DCT coding, it was shown that the above approximation yields a higher experimental SNR and a better agreement between theory and simulation than the Gaussian or Laplacian assumptions.

• Kyungpook Mathematical Journal
• /
• v.57 no.3
• /
• pp.493-502
• /
• 2017

We propose an accurate approximation method via discrete Krawtchouk orthogonal polynomials to the distribution of a sum of independent but non-identically distributed binomial random variables. This approximation is a weighted binomial distribution with no need for continuity correction unlike commonly used density approximation methods such as saddlepoint, Gram-Charlier A type(GC), and Gaussian approximation methods. The accuracy obtained from the proposed approximation is compared with saddlepoint approximations applied by Eisinga et al. [4], which are the most accurate method among higher order asymptotic approximation methods. The numerical results show that the proposed approximation in general provide more accurate estimates over the entire range for the target probability mass function including the right-tail probabilities. In addition, the method is mathematically tractable and computationally easy to program.

• Communications for Statistical Applications and Methods
• /
• v.20 no.1
• /
• pp.15-22
• /
• 2013

We find a sufficient condition for optimal Berry-Esseen bounds in the normal approximation of functionals of Gaussian fields studied by Nourdin and Peccati (2009b).