• Communications for Statistical Applications and Methods
• /
• v.17 no.1
• /
• pp.17-27
• /
• 2010

Feature selection and feature weighting are useful techniques for improving the classification accuracy of k-Nearest Neighbor (k-NN) classifier. The main propose of feature selection and feature weighting is to reduce the number of features, by eliminating irrelevant and redundant features, while simultaneously maintaining or enhancing classification accuracy. In this paper, a novel hybrid approach is proposed for simultaneous feature selection, feature weighting and choice of k in k-NN classifier based on Genetic Algorithm. The results have indicated that the proposed algorithm is quite comparable with and superior to existing classifiers with or without feature selection and feature weighting capability.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.4E
• /
• pp.31-38
• /
• 2001

A large-scale synthesis database for a unit selection based synthesis method usually retains redundant synthesis unit instances, which are useless to the synthetic speech quality. In this paper, to eliminate those instances from the synthesis database, we proposed a new pruning method called weighted vector quantization (WVQ). The WVQ reflects relative importance of each synthesis unit instance when clustering the similar instances using vector quantization (VQ) technique. The proposed method was compared with two conventional pruning methods through the objective and subjective evaluations of the synthetic speech quality: one to simply limit maximum number of instance, and the other based on normal VQ-based clustering. The proposed method showed the best performance under 50% reduction rates. Over 50% of reduction rates, the synthetic speech quality is not seriously but perceptibly degraded. Using the proposed method, the synthesis database can be efficiently reduced without serious degradation of the synthetic speech quality.

• Proceedings of the KIEE Conference
• /
• 1999.07b
• /
• pp.731-733
• /
• 1999

Distributed Control System has been used for large scale and critical system control such as aerospace industries, chemical and power plant and so on. It is very impotant factors for design of the control system to be reliable and fault-tolerant. These control systems have backup or redundant processing modules for minimizing the time of failure and improving reliability. But such methods have changeover duration from faulty module to healthy one. During that interval, feedback control loop raises bumper and performance of the system become worse. TMR(Triple Modular Redundancy) control system is one of the best reliable ones that can overcome such a mortal drawback. This paper analyzes the components of TMR system functionally and proposes practical and cost effective configuration method for turbine control of thermal power plant.

• Proceedings of the Korea Concrete Institute Conference
• /
• 1993.10a
• /
• pp.195-200
• /
• 1993

This paper develops practical and realistic reliability models and methods for the evalusion of system reliability and system reliability-based rating of R.C box-girder bridge superstructures. The precise prediction of reserved carrying capacity of bridge as a system is extremely difficult expecially when the bridges are highly redundant and significantly deteriorated or damaged. This paper proposes a new approach for the evaluation of reserved system carrying capacity of bridges in terms of equivalent system-strength, which may be defined as a bridge system-strength corresponding to the system reliability of the bridge. This can be derived from an inverse process based on the concept of FOSM form of system reliability index. The strength limit state models for R.C box-girder bridges suggested in the paper are based on the basic bending and shear strength. and the system reliability problem of box-girder superstructure is formulated as parallel-series models obtained from the FMA(Failure Mode Approach) based on major failure mechanism or critical failure states of each girder. AFOSM and IST(Importance Sampling Technique) simulation algorithm is used for the reliability analysis of the proposed models.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.9
• /
• pp.133-141
• /
• 1994

Conventional competitive learning algorithms compute the Euclidien distance to determine the winner neuron out of all predetermined output neurons. In such cases, there is a drawback that the performence of the learning algorithm depends on the initial reference(=weight) vectors. In this paper, we propose a new competitive learning algorithm that dynamically generates output neurons. The proposed method generates output neurons by dynamically changing the class thresholds for all output neurons. We compute the similarity between the input vector and the reference vector of each output neuron generated. If the two are similar, the reference vector is adjusted to make it still more like the input vector. Otherwise, the input vector is designated as the reference vector of a new outputneuron. Since the reference vectors of output neurons are dynamically assigned according to input pattern distribution, the proposed method gets around the phenomenon that learning is early determined due to redundant output neurons. Experiments using speech data have shown the proposed method to be superior to existint methods.

• Communications for Statistical Applications and Methods
• /
• v.16 no.4
• /
• pp.573-578
• /
• 2009

We prove the consistency of a penalized maximum likelihood estimate proposed by Ahn (2001). The PMLE not only avoids the well-known problem that the ordinary likelihood of the normal mixture model is unbounded for any given sample size, but also removes redundant components.

• Journal of Mechanical Science and Technology
• /
• v.19 no.spc1
• /
• pp.237-254
• /
• 2005

The vast mechanical systems could be classified as closed loop system, open loop system and open & closed (switching) system. In the closed loop system, the kinematics and dynamics of 3-D mechanisms will be reviewed and closed form solutions using the direction cosine matrix method and reflection transformation method will be introduced. In the open loop system, kinematic & dynamic analysis methods regarding the redundant system which has more degrees of freedom in joint space than those of task space are reviewed and discussed. Finally, switching system which changes its phase between closed and open loop motion is investigated with the principle of dynamical balance. Among switching systems, the human gait in biomechanics and humanoid in robotics are presented.

• Journal of Biosystems Engineering
• /
• v.25 no.2
• /
• pp.141-150
• /
• 2000

This study was carried out to de develop a control strategy of a fruit harvesting redundant robot. The method of generating a safe trajectory, which avoids collisions with obstracles such as branches or immature fruits, in the 3D(3-dimension) space using artificial potential field technique and virtual plane concept was proposed. Also, the method of setting reference velocity vectors to follow the trajectory and to avoid obstacles in the 3D space was proposed. Developed methods were verified with computer simulations and with actual robot tests. Fro the actual robot tests, a machine vision system was used for detecting fruits and obstacles, Results showed that developed control method could reduce the occurrences of the robot manipulator located in the possible collision distance. with 10 virtual obstacles generated randomly in the 3 D space, maximum rates of the occurrences of the robot manipulator located in the possible collision distance, 0.03 m, from the obstacles were 8 % with 5 degree of freedom (DOF), 8 % with 6-DOF, and 4% with 7-DOF, respectively.

• Korean Management Science Review
• /
• v.23 no.1
• /
• pp.63-74
• /
• 2006

Decision tree induction algorithm is one of the most widely used methods in classification problems. However, they could be trapped into a local minimum and have no reasonable means to escape from it if tree algorithm uses top-down search algorithm. Further, if irrelevant or redundant features are included in the data set, tree algorithms produces trees that are less accurate than those from the data set with only relevant features. We propose a hybrid algorithm to generate decision tree that uses genetic programming with sequentially selected features. Correlation-based Feature Selection (CFS) method is adopted to find relevant features which are fed to genetic programming sequentially to find optimal trees at each iteration. The new proposed algorithm produce simpler and more understandable decision trees as compared with other decision trees and it is also effective in producing similar or better trees with relatively smaller set of features in the view of cross-validation accuracy.

• ETRI Journal
• /
• v.41 no.6
• /
• pp.750-759
• /
• 2019

Presently, the extraction of hand-crafted features is still the dominant method in radar emitter recognition. To solve the complicated problems of selection and updation of empirical features, we present a novel automatic feature extraction structure based on deep learning. In particular, a convolutional neural network (CNN) is adopted to extract high-level abstract representations from the time-frequency images of emitter signals. Thus, the redundant process of designing discriminative features can be avoided. Furthermore, to address the performance degradation of a single platform, we propose the construction of an ensemble learning-based architecture for multi-platform fusion recognition. Experimental results indicate that the proposed algorithms are feasible and effective, and they outperform other typical feature extraction and fusion recognition methods in terms of accuracy. Moreover, the proposed structure could be extended to other prevalent ensemble learning alternatives.