• Journal of the Korean Operations Research and Management Science Society
• /
• v.30 no.1
• /
• pp.161-176
• /
• 2005

In a situation that ordinal preferences on multiattribute weights are captured, we present two solution approaches: an exact approach and an approximate method. The former, an exact solution approach via interaction with a decision-maker, pursues the progressive reduction of a set of non-dominated alternatives by narrowing down the feasible attribute weights region. Subsequent interactive questions and responses, however, sometimes may not guarantee the best alternative or a complete rank order of a set of alternatives that the decision-maker desires to have. Approximate solution approaches, on the other hand, can be divided into three categories including surrogate weights methods, dominance value-based decision rules, and three classical decision rules. Their efficacies are evaluated in terms of choice accuracy via a simulation analysis. The simulation results indicate that a proposed hybrid approach, intended to combine an exact solution approach through interaction and a dominance value-based approach, is recommendable for aiding a decision making in a case that a final choice is seldom made at single step under attribute weights that are imprecisely specified beyond ordinal descriptions.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2004.10a
• /
• pp.143-146
• /
• 2004

In a situation that rank order information on attribute weights is captured, two solution approaches are presented. An exact solution approach via interaction with a decision-maker pursues progressive reduction of a set of non-dominated alternatives by narrowing down the feasible attribute weights set. In approximate solution approach, on the other hand, three categories of approximate methods such as surrogate weights method, the dominance value-based decision rules, and three classical decision rules are presented and their efficacies in terms of choice accuracy are evaluated via simulation analysis. The simulation results indicate that a method, which combines an exact solution approach through interactions with the decision-maker and the dominance value-based approach is recommendable in a case that a decision is not made at a single step under imprecisely assessed weights information.

• Korean Journal of Mathematics
• /
• v.23 no.3
• /
• pp.327-336
• /
• 2015

In this paper, we study a constructive approximation by neural networks with positive integer weights. Like neural networks with real weights, we show that neural networks with positive integer weights can even approximate arbitrarily well for any continuous functions on compact subsets of $\mathbb{R}$. We give a numerical result to justify our theoretical result.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.19 no.2
• /
• pp.273-278
• /
• 2009

The determination of the associated weights in the theory of ordered weighted averaging (OWA) operators is one of the important issue. Recently, Wang and Parkan [Information Sciences 175 (2005) 20-29] proposed a minimax disparity approach for obtaining OWA operator weights and the approach is based on the solution of a linear program (LP) model for a given degree of orness. Recently, Liu [International Journal of Approximate Reasoning, accepted] showed that the minimum variance OWA problem of Fuller and Majlender [Fuzzy Sets and Systems 136 (2003) 203-215] and the minimax disparity OWA problem of Wang and Parkan always produce the same weight vector using the dual theory of linear programming. In this paper, we give an improved proof of the minimax disparity problem of Wang and Parkan while Liu's method is rather complicated. Our method gives the exact optimum solution of OWA operator weights for all levels of orness, $0\leq\alpha\leq1$, whose values are piecewise linear and continuous functions of $\alpha$.

• Journal of the Korean Data and Information Science Society
• /
• v.16 no.4
• /
• pp.1045-1052
• /
• 2005

This paper deals with the estimations of kernel ridge regression when the responses are subject to randomly right censoring. The weighted data are formed by redistributing the weights of the censored data to the uncensored data. Then kernel ridge regression can be taken up with the weighted data. The hyperparameters of model which affect the performance of the proposed procedure are selected by a generalized approximate cross validation(GACV) function. Experimental results are then presented which indicate the performance of the proposed procedure.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.41.5-41
• /
• 2001

A novel method is introduced for determining the weights of a regularization network which approximates one-to-many mapping. A conventional neural network will converges to the average value when outputs are multiple for one input. The capability of proposed network is demonstrated by an example of learning inverse mapping.

• Smart Structures and Systems
• /
• v.5 no.6
• /
• pp.613-632
• /
• 2009

The present paper is concerned with the design of a proper patch actuator network in order to track a desired displacement of the sidewalls of a one-storey frame structure; both, for the static and the dynamic case. Weights for each patch of the actuator network found in our previous work were based on beam theory; in the present paper a refinement of these weights by modeling the sidewalls of the frame structure as thin plates is presented. For the sake of calculating the refined weights approximate solutions of the plate equations are calculated by an extended Galerkin method. The solutions based on the analytical plate model are compared with three-dimensional Finite Element results computed in the commercially available code ANSYS. The patch actuator network is put into practice by means of four piezoelectric patches attached to each of the two sidewalls of the frame structures, to which electric voltages proportional to the analytically refined patch weights are applied. Analytical and numerical results coincide very well over a broad frequency range.

• Toxicological Research
• /
• v.26 no.1
• /
• pp.53-59
• /
• 2010

This study was conducted to obtain acute information of the oral dose toxicity of DHU001, a polyherbal formula in male and female mice. In order to calculated 50% lethal dose ($LD_{50}$) and approximate lethal dose (LD), test material was once orally administered to male and female ICR mice at dose levels of 2000, 1000, 500, 250 and 0 (vehicle control) ml/kg (body weight). The mortality and changes on body weight, clinical signs, gross observation, organ weight and histopathology of principle organs were monitored 14 days after treatment with DHU001. We could not find any mortalities, DHU001 treatment-related clinical signs, changes on the body and organ weights, gross and histopathological findings. The results obtained in this study suggest that $LD_{50}$ and approximate LD in mice after single oral dose of DHU001 were considered over 2000 mg/kg in both female and male mice.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.7
• /
• pp.32-39
• /
• 1996

In this paper, we describe the algorithm which controls an unknown nonlinear system with multilayer neural network. The multilayer neural netowrk can be used to approximate any continuous function to any desired degree of accuracy. With the former fact, we approximate unknown nonlinear function on the nonlinear system by using of multilayer neural netowrk. The weights on the hidden layer of multilayer neural network are updated by gradient method. The weight-update rule on the output layer is derived to satisfy lyapunov stability. Also, we obtain secondary controller form deriving step. The global control system consists of controller using feedback linearization method and secondary controller is order to satisfy layapunov stability. The proposed control algorithm is verified through computer simulation.

• Transactions on Control, Automation and Systems Engineering
• /
• v.2 no.3
• /
• pp.189-195
• /
• 2000

The mathematical solutions of the stability convergence are important problems in system control. In this paper such problems are analyzed and resolved for system control using multilayer neural networks. We describe an algorithm to control an unknown nonlinear system with a disturbance, using a multilayer neural network. We include a disturbance among the modeling error, and the weight update rules of multilayer neural network are derived to satisfy Lyapunov stability. The overall control system is based upon the feedback linearization method. The weights of the neural network used to approximate a nonlinear function are updated by rules derived in this paper . The proposed control algorithm is verified through computer simulation. That is as the weights of neural network are updated at every sampling time, we show that the output error become finite within a relatively short time.