• Journal of KIISE:Software and Applications
• /
• v.28 no.9
• /
• pp.603-611
• /
• 2001

This paper proposes a novel fast layer-by-layer algorithm that has better generalization capability. In the proposed algorithm, the weights of the hidden layer are updated by the target vector of the hidden layer obtained by least squares method. The proposed algorithm improves the learning speed that can occur due to the small magnitude of the gradient vector in the hidden layer. This algorithm was tested in a handwritten digits recognition problem. The learning speed of the proposed algorithm was faster than those of error back propagation algorithm and modified error function algorithm, and similar to those of Ooyen's method and layer-by-layer algorithm. Moreover, the simulation results showed that the proposed algorithm had the best generalization capability among them regardless of the number of hidden nodes. The proposed algorithm has the advantages of the learning speed of layer-by-layer algorithm and the generalization capability of error back propagation algorithm and modified error function algorithm.

• Communications for Statistical Applications and Methods
• /
• v.26 no.5
• /
• pp.473-495
• /
• 2019

In this paper, a new extension of Lindley distribution has been introduced. Certain characterizations based on truncated moments, hazard and reverse hazard function, conditional expectation of the proposed distribution are presented. Besides, these characterizations, other statistical/mathematical properties of the proposed model are also discussed. The estimation of the parameters is performed through different classical methods of estimation. Bayes estimation is computed under gamma informative prior under the squared error loss function. The performances of all estimation methods are studied via Monte Carlo simulations in mean square error sense. The potential of the proposed model is analyzed through two data sets. A modified goodness-of-fit test using the Nikulin-Rao-Robson statistic test is investigated via two examples and is observed that the new extension might be used as an alternative lifetime model.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1810-1815
• /
• 2003

Sliding mode control with nonlinear interpolation in the boundary layer is proposed. A modified sigmoid function is used for nonlinear interpolation in the boundary layer and its parameter is tuned by a fuzzy logic controller. The fuzzy logic controller that takes the distance between the system state and the sliding surface as its input guides the choice of parameter of the modified sigmoid function and the parameter is on-line tuned. Owing to the decreased thickness, the proposed method has better tracking performance than the conventional linear interpolation method. To demonstrate its performance, the proposed control algorithm is applied to a simple nonlinear system model.

• International Journal of Control, Automation, and Systems
• /
• v.2 no.4
• /
• pp.523-529
• /
• 2004

Sliding mode control with nonlinear interpolation in the boundary layer is proposed. A modified sigmoid function is used for nonlinear interpolation in the boundary layer and its parameter is tuned by a fuzzy controller. The fuzzy controller that takes both the sliding variable and a measure of chattering as its inputs tunes the parameter of the modified sigmoid function. Owing to the decreased thickness of the boundary layer and the tuned parameter, the proposed method has superior tracking performance than the conventional linear interpolation method.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.30 no.8 s.251
• /
• pp.923-930
• /
• 2006

A Kriging model is a sort of approximation model and used as a deterministic model of a computationally expensive analysis or simulation. Although it has various advantages, it is difficult to assess the accuracy of the approximated model. It is generally known that a mean square error (MSE) obtained from the kriging model can't calculate statistically exact error bounds contrary to a response surface method, and a cross validation is mainly used. But the cross validation also has many uncertainties. Moreover, the cross validation can't be used when a maximum error is required in the given region. For solving this problem, we first proposed a modified mean square error which can consider relative errors. Using the modified mean square error, we developed the strategy of adding a new sample to the place that the MSE has the maximum when the MSE is used for the assessment of the kriging model. Finally, we offer guidelines for the use of the MSE which is obtained from the kriging model. Four test problems show that the proposed strategy is a proper method which can assess the accuracy of the kriging model. Based on the results of four test problems, a convergence coefficient of 0.01 is recommended for an exact function approximation.

• Journal of KSNVE
• /
• v.8 no.5
• /
• pp.841-848
• /
• 1998

In the gear manufacturing, tooth modification is usually applied for the prevention of tooth impact during the loading. In contrary, tooth profile error causes amplifying the whine noise which is cumbersome to reduce in the automobile gear box. So optimum quantity of the modifications must be obtained for the good performance in the vibrational sense. In this paper, a formulation to define the tooth curve by considering the profile manufacturing error and loading deformation of the gear tooth is suggested and the transmission error and loading deformation of the gear tooth is suggested and the transmission error with modified tooth in the gear system is evaluated. A pair of gear set is mathematically modelled. The equivalent excitation in the gear vibratonal model is formulated. For the experimental evaluaton on the derived transmission error function, a simple geared system is set up in which the gears are designed to give pre-designed tooth profile modification and manufactured by CNC Wire Cutting Machine. Under slow speed operaton, the transmission error of the gear pair is measured by using two rotational laser vibrometers, compared with the calculated one of which the result shows good agreement.

• Speech Sciences
• /
• v.14 no.1
• /
• pp.163-174
• /
• 2007

All observation has uncertainty due to noise or channel characteristics. This uncertainty should be counted in the modeling of observation. In this paper we propose a modified optimization object function of a GMM training considering inexact observation. The object function is modified by introducing the concept of observation confidence as a weighting factor of probabilities. The optimization of the proposed criterion is solved using a common EM algorithm. To verify the proposed method we apply it to the speaker recognition domain. The experimental results of text-independent speaker identification with VidTimit DB show that the error rate is reduced from 14.8% to 11.7% by the modified GMM training.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.627-630
• /
• 1999

For depth extraction from the focus and recovery the shape, determination of criterion function for focus measure and size of the criterion window are very important. However, Texture, illumination, and magnification have an effect on focus measure. For that reason, depth map has a partial high and low peak. In this paper, we propose a depth extraction method from focused images using the error interpolation. This method is modified the error depth into mean value between two normal depth in order to improve the depth map.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.12
• /
• pp.2498-2504
• /
• 2009

The fundamental research for the mobile robot navigation using the numerical optimization method is presented. We propose an image-based visual servo navigation algorithm for a wheeled mobile robot utilizing a ceiling mounted camera. For the image-based visual servoing, we define the composite image Jacobian which represents the relationship between the speed of wheels of a mobile robot and the robot's overall speed in the image plane. The rotational speed of wheels of a mobile robot can be directly related to the overall speed of a mobile robot in the image plane using the composite image Jacobian. We define the mobile robot navigation problem as an unconstrained optimization problem to minimize the cost function with the image error between the goal position and the position of a mobile robot. In order to avoid the obstacle, the modified cost function is proposed which is composed of the image error between the position of a mobile robot and the goal position and the distance between the position of a mobile robot and the position of the obstacle. The performance was evaluated using the simulation.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.12 no.5
• /
• pp.107-114
• /
• 2012

In this paper, the concerned existing blind equalizer convergence rate and residual inter-symbol interference using constellation reduced and cost function by separation the real part and an imaginary part, the dual structure CR-CMA(constellation Reduction CMA). The CMA methed compensates amplitude but does no compensate phase, On the other hand, The CMA method compensates both the amplitude and the phase but it has the convergence rate problem, and the MCMA method is a way to solve the phase problem of CMA method compensates both the amplitude and the phase after respectively calculating the real part and imaginary part components. Proposal a new method that the dual structure of CR-CMA, the cost function and error function and respectively calculating the real part and imaginary part components can advantages by improving the CMA and the MCMA algorithms so that the amplitude and phase retrieval and constellation reduce the residual ISI and faster convergence rate and performance is good SER (Symbol Error Ratio) was confirmed by computer simulations.