• Journal of the Korean Data and Information Science Society
• /
• v.23 no.6
• /
• pp.1231-1239
• /
• 2012

An estimating procedure is introduced for kernel Poisson regression when the input variables consist of numerical and categorical variables, which is based on the penalized negative log-likelihood and the component-wise product of two different types of kernel functions. The proposed procedure provides the estimates of the mean function of the response variables, where the canonical parameter is linearly and/or nonlinearly related to the input variables. Experimental results are then presented which indicate the performance of the proposed kernel Poisson regression.

• Journal of the Korean Data and Information Science Society
• /
• v.18 no.3
• /
• pp.767-772
• /
• 2007

A kernel machine is proposed as an estimating procedure for the linear and nonlinear Poisson regression, which is based on the penalized negative log-likelihood. The proposed kernel machine provides the estimate of the mean function of the response variable, where the canonical parameter is related to the input vector in a nonlinear form. The generalized cross validation(GCV) function of MSE-type is introduced to determine hyperparameters which affect the performance of the machine. Experimental results are then presented which indicate the performance of the proposed machine.

• Journal of the Korean Data and Information Science Society
• /
• v.19 no.4
• /
• pp.1353-1360
• /
• 2008

An estimating procedure is introduced for the nonlinear mixed-effect Poisson regression, for longitudinal study, where data from different subjects are independent whereas data from same subject are correlated. The proposed procedure provides the estimates of the mean function of the response variables, where the canonical parameter is related to the input vector in a nonlinear form. The generalized cross validation function is introduced to choose optimal hyper-parameters in the procedure. Experimental results are then presented, which indicate the performance of the proposed estimating procedure.

• Journal of Mechanical Science and Technology
• /
• v.17 no.10
• /
• pp.1443-1449
• /
• 2003

Primary goal of adaptive observers would be to estimate the true states of a plant. Identification of unknown parameters is of secondary interest and is achieved frequently with the persistent excitation condition of some regressors. Nevertheless, two problems are linked to each other in the classical approaches to adaptive observers; as a result, we get a good state estimate once after a good parameter estimate is obtained. This paper focuses on the state estimation without parameter identification so that the state is estimated regardless of persistent excitation. In this direction of research, Besancon (2000) recently summarized that most of adaptive observers in the literature share one common canonical form, in which unknown parameters do not affect the unmeasured states. We enlarge the class of linear systems from the canonical form of (Besancon, 2000) by proposing an adaptive observer (with additional dynamics) that allows unknown parameters to affect those unmeasured states. A recursive algorithm is presented to design the proposed dynamic observer systematically. An example confirms the design procedure with a simulation result.

• 전기의세계
• /
• v.31 no.5
• /
• pp.361-367
• /
• 1982

The problem of parameter identification for multivariable stochastic linear systems from output measurements, which are corrupted by noises, is considered. A modified Luenberger's input/output canonical form is used for reducing the number of unknown coefficients. A computationally and conceptionally simple systematic procedure for parameter estimation is obtained using output correlation method. The estimates are shown to be asymptotically normal, unbiased and consistent. Numerical examples are presented to illustrate the identification method.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.12 no.5
• /
• pp.481-486
• /
• 2002

In this paper, a new on-line parameter estimation methodology for the general continuous time Takagi-Sugeno(T-5) fuzzy model whose parameters are poorly known or uncertain is presented. An estimator with an appropriate adaptive law for updating the parameters is designed and analyzed based on the Lyapunov theory. The adaptive law is designed so that the estimation model follows the plant parameterized model. By the proposed estimator, the parameters of the T-S fuzzy model can be estimated by observing the behavior of the system and it can be a basis for the indirect adaptive fuzzy control. Based on the derived design method, the parameter estimation for controllable canonical T-S fuzzy model is also Presented.

• Journal of electromagnetic engineering and science
• /
• v.13 no.1
• /
• pp.38-43
• /
• 2013

We derive the Z-parameters for the two coupled antennas used for wireless power transfer under the assumption that the antennas are canonical minimum scattering antennas. Using the Z-parameter and the maximum power transfer efficiency formula, we determine the maximum power transfer efficiency of wireless power transfer systems. The results showed that the maximum power transfer efficiency increases as the mode number or the radiation efficiency increases. To verify the theory, we fabricate and measure two different power transfer systems: one comprises two antennas generating $TM_{01}$ mode; the other comprises two antennas generating $TM_{02}$ mode. When the distance between the centers of the antennas was 30 cm, the maximum power transfer efficiency of the antennas generating the $TM_{02}$ mode increased by 62 % compared to that of the antennas generating the $TM_{01}$ mode.

• Proceedings of the KIEE Conference
• /
• 1996.07b
• /
• pp.1060-1062
• /
• 1996

This paper presents a discrete adaptive observer for MIMO system of an IM model in DQ reference model. The IM model in the stationary frame is discretized and it is transformed into the canonical observer form. The unknown parameter is choosen as rotor speed. The adaptive law for parameter adjustment is obtained as a set of recursive equations which are derived by utilizing an exponentially weighted normalized least-square method. The proposed adaptive observer converges rapidly and is also shown to track time-varying plant parameter quickly. Its effectiveness has been demonstrated by computer simulation.

• Journal of the Korean Mathematical Society
• /
• v.53 no.4
• /
• pp.869-893
• /
• 2016

We construct the extension of a hyperelliptic K3 surface to a Fano 6-fold with extraordinary properties in moduli. This leads us to a family of surfaces of general type with $p_g=1$, q = 0, $K^2=2$ and hyperelliptic canonical curve, each of which is a weighted complete inter-section inside a Fano 6-fold. Finally, we use these hyperelliptic surfaces to determine an 8-parameter family of Godeaux surfaces with ${\pi}_1={\mathbb{Z}}/2$.

• Proceedings of the IEEK Conference
• /
• 2002.07b
• /
• pp.810-813
• /
• 2002

We present the information theoretic learning based on the Tsallis entropy maximization principle for various q. The Tsallis entropy is one of the generalized entropies and is a canonical entropy in the sense of physics. Further, we consider the dependency of the learning on the parameter $\sigma$, which is a standard deviation of an assumed a priori distribution of samples such as Parzen window.