• 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.

• Journal of Korean Society on Water Environment
• /
• v.23 no.4
• /
• pp.551-560
• /
• 2007

The present study analyzed noise reduction and long/short-term components for discharge, TOC concentration, and TOC load data in order to understand the data characteristics better. For the purpose, wavelet transform which can reduce noise from raw data and has flexible resolution in time and frequency domain was applied and the theory of nonlinear dynamics was also used to determine the last decomposition level for wavelet transform. Wavelet function of 'db10' and the 7th level for the last decomposition of wavelet transform were applied for the all data in the present study. Also the results revealed that the energy ratios of approximation components with 187-hour periodicity decomposed from 7th level of wavelet transform were 94.71% (discharge), 99.00% (TOC concentration), and 93.84% (TOC load), respectively. In addition, the energy ratios of detail components showed the range between 1.00% and 6.17%, which were extremely small comparing to the energy ratios of approximation components, therefore, the first and second detail components might be considered as noise components included in the raw data.

• Proceedings of the KIEE Conference
• /
• 2005.05a
• /
• pp.106-108
• /
• 2005

The calculation of parameters of a process model is modified to find that better sliding mode controller for a process with time delay. A design method by Camacho has such problems as chattering, overshoot due to the Pade approximation errors for the time delay term of the first order model. In this paper, a new design technique for a sliding mode controller is proposed by introducing the modified Pade approximation considering the weight factor. The modified method is expected to provide a more accurate model with better control settings.

• Steel and Composite Structures
• /
• v.20 no.5
• /
• pp.1087-1102
• /
• 2016

The paper proposes to characterize the free vibration behaviour of non-uniform cylindrical shells using spline approximation under first order shear deformation theory. The system of coupled differential equations in terms of displacement and rotational functions are obtained. These functions are approximated by cubic splines. A generalized eigenvalue problem is obtained and solved numerically for an eigenfrequency parameter and an associated eigenvector which are spline coefficients. Four and two layered cylindrical shells consisting of two different lamination materials and plies comprising of same as well as different materials under two different boundary conditions are analyzed. The effect of length parameter, circumferential node number, material properties, ply orientation, number of lay ups, and coefficients of thickness variations on the frequency parameter is investigated.

• Proceedings of the IEEK Conference
• /
• 2002.06e
• /
• pp.47-50
• /
• 2002

An extended-counting analog to digital converter (ADC) is designed to have a high resolution(14bit) with low power consumption and small dia area. First order sigma-delta modulator with a simple counter for incremental operation eliminates the need of big decimation filter in conventional sigma-delta type ADC. To improve the accuracy and linearity, extended mode of successive approximation is followed. For 14-bit conversion operation, total 263 clocks(1 clock for reset, 256 clocks for incremental operation and extended 6 clocks for successive approximation operation) are needed with the sampling rate of 10 Ms/s This ADC is implemented in a 0.6um standard CMOS technology with a die area of 1 mm ${\times}$ 0.75 mm.

• East Asian mathematical journal
• /
• v.27 no.5
• /
• pp.545-555
• /
• 2011

Based on the A-maximal(m)-relaxed monotonicity frameworks, the approximation solvability of a general class of variational inclusion problems using the relaxed proximal point algorithm is explored, while generalizing most of the investigations, especially of Xu (2002) on strong convergence of modified version of the relaxed proximal point algorithm, Eckstein and Bertsekas (1992) on weak convergence using the relaxed proximal point algorithm to the context of the Douglas-Rachford splitting method, and Rockafellar (1976) on weak as well as strong convergence results on proximal point algorithms in real Hilbert space settings. Furthermore, the main result has been applied to the context of the H-maximal monotonicity frameworks for solving a general class of variational inclusion problems. It seems the obtained results can be used to generalize the Yosida approximation that, in turn, can be applied to first- order evolution inclusions, and can also be applied to Douglas-Rachford splitting methods for finding the zero of the sum of two A-maximal (m)-relaxed monotone mappings.

• Communications for Statistical Applications and Methods
• /
• v.29 no.4
• /
• pp.471-486
• /
• 2022

Autoregressive moving average (ARMA) models involve nonlinearity in the model coefficients because of unobserved lagged errors, which complicates the likelihood function and makes the posterior density analytically intractable. In order to overcome this problem of posterior analysis, some approximation methods have been proposed in literature. In this paper we first review the main analytic approximations proposed to approximate the posterior density of ARMA models to be analytically tractable, which include Newbold, Zellner-Reynolds, and Broemeling-Shaarawy approximations. We then use the Kullback-Leibler divergence to study the relation between these three analytic approximations and to measure the distance between their derived approximate posteriors for ARMA models. In addition, we evaluate the impact of the approximate posteriors distance in Bayesian estimates of mean and precision of the model coefficients by generating a large number of Monte Carlo simulations from the approximate posteriors. Simulation study results show that the approximate posteriors of Newbold and Zellner-Reynolds are very close to each other, and their estimates have higher precision compared to those of Broemeling-Shaarawy approximation. Same results are obtained from the application to real-world time series datasets.

• Wind and Structures
• /
• v.35 no.5
• /
• pp.323-336
• /
• 2022

This study investigates the applicability of the direct identification of flutter derivatives in the time domain using Rational Function Approximation (RFA), where the extraction procedure requires either a combination of at least two wind speeds or one wind speed. In the frequency domain, flutter derivatives are identified at every wind speed. The ease of identifying flutter derivatives in the time domain creates a paradox because flutter derivative patterns sometimes change in higher-order polynomials. The first step involves a numerical study of RFA extractions for different deck shapes from existing bridges to verify the accurate wind speed combination for the extraction. The second step involves validating numerical simulation results through a wind tunnel experiment using the forced vibration method in one degree of freedom. The findings of the RFA extraction are compared to those obtained using the analytical solution. The numerical study and the wind tunnel experiment results are in good agreement. The results show that the evolution pattern of flutter derivatives determines the accuracy of the direct identification of RFA.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.27 no.1
• /
• pp.37-55
• /
• 2023

In this paper we propose a new algorithm for detecting and counting flowers in a complex background based on digital images. The algorithm mainly includes the following parts: edge contour extraction of flowers, edge contour determination of overlapped flowers and flower counting. We use a contour detection technique in Computer Vision (CV) to extract the edge contours of flowers and propose an improved algorithm with a concave point detection technique to find accurate segmentation for overlapped flowers. In this process, we first use the polygon approximation to smooth edge contours and then adopt the second-order central moments to fit ellipse contours to determine whether edge contours overlap. To obtain accurate segmentation points, we calculate the curvature of each pixel point on the edge contours with an improved Curvature Scale Space (CSS) corner detector. Finally, we successively give three adaptive judgment criteria to detect and count flowers accurately and automatically. Both experimental results and the proposed evaluation indicators reveal that the proposed algorithm is more efficient for flower counting.

• Environmental Engineering Research
• /
• v.24 no.2
• /
• pp.309-317
• /
• 2019

This study evaluated the kinetics of acrylamide (AM) biodegradation by mixed culture bacteria and Enterobacter aerogenes (E. aerogenes) in sequencing batch reactor (SBR) systems with AQUASIM and linear regression. The zero-order, first-order, and Monod kinetic models were used to evaluate the kinetic parameters of both autotrophic and heterotrophic nitrifications and both AM and chemical oxygen demand (COD) removals at different AM concentrations of 100, 200, 300, and 400 mg AM/L. The results revealed that both autotrophic and heterotrophic nitrifications and both AM and COD removals followed the Monod kinetics. High AM loadings resulted in the transformation of Monod kinetics to the first-order reaction for AM and COD removals as the results of the compositions of mixed substrates and the inhibition of the free ammonia nitrogen (FAN). The kinetic parameters indicated that E. aerogenes degraded AM and COD at higher rates than mixed culture bacteria. The FAN from the AM biodegradation increased both heterotrophic and autotrophic nitrification rates at the AM concentrations of 100-300 mg AM/L. At higher AM concentrations, the FAN accumulated in the SBR system inhibited the autotrophic nitrification of mixed culture bacteria. The accumulation of intracellular polyphosphate caused the heterotrophic nitrification of E. aerogenes to follow the first-order approximation.