• 한국공간정보시스템학회:학술대회논문집
• /
• 2002.03a
• /
• pp.45-52
• /
• 2002

Artificial neural networks (ANN) have been successfully used for classifying remotely sensed imagery. However, ANN still is not the preferable choice for classification over the conventional classification methodology such as the maximum likelihood classifier commonly used in the industry production environment. This can be attributed to the ANN characteristic built-in stochastic process that creates difficulties in dealing with unequally represented training classes, and its training performance speed. In this paper we examined some practical aspects of training classes when using a back propagation neural network model for remotely sensed imagery. During the classification process of remotely sensed imagery, representative training patterns for each class are collected by polygons or by using a region-growing methodology over the imagery. The number of collected training patterns for each class may vary from several pixels to thousands. This unequally populated training data may cause the significant problems some neural network empirical models such as back-propagation have experienced. We investigate the effects of training over- or under- represented training patterns in classes and propose the pattern repopulation algorithm, and an adaptive alpha adjustment (AAA) algorithm to handle unequally represented classes. We also show the performance improvement when input patterns are presented in random fashion during the back-propagation training.

• Journal of Digital Convergence
• /
• v.12 no.3
• /
• pp.249-254
• /
• 2014

Wireless LAN system is evolving toward high-speed data transmission and more accurate channel estimation is necessarily required to improve communication performance. The PLCP preamble field in IEEE 802.11 based wireless MODEM consists of ten short symbols and two long symbols and is used for synchronization and channel estimation. The existing least square (LS) channel estimation is based on only two long training symbols. After estimating channel response separately by using each long training symbol, the final channel estimation is obtained by the average of each estimation. In this paper, a new channel estimation algorithm is presented to improve the performance of the existing LS channel estimation algorithm. From the fact that the short training symbol consists of 12 non-zero subcarriers, it gives us a clue of being able to additionally estimate at least one fourth of channel coefficients. The new LS algorithm performs channel estimation based on both two long training symbols and a short training symbol. The proposed LS algorithm shows a little bit performance improvement over the existing LS estimation and it will be able to be applied to the IEEE 802.11p WAVE system.

• Journal of information and communication convergence engineering
• /
• v.13 no.2
• /
• pp.123-131
• /
• 2015

Training neural networks is a complex task with great importance in the field of supervised learning. In the training process, a set of input-output patterns is repeated to an artificial neural network (ANN). From those patterns weights of all the interconnections between neurons are adjusted until the specified input yields the desired output. In this paper, a new hybrid algorithm is proposed for global optimization of connection weights in an ANN. Dynamic swarms are shown to converge rapidly during the initial stages of a global search, but around the global optimum, the search process becomes very slow. In contrast, the gradient descent method can achieve faster convergence speed around the global optimum, and at the same time, the convergence accuracy can be relatively high. Therefore, the proposed hybrid algorithm combines the dynamic particle swarm optimization (DPSO) algorithm with the backpropagation (BP) algorithm, also referred to as the DPSO-BP algorithm, to train the weights of an ANN. In this paper, we intend to show the superiority (time performance and quality of solution) of the proposed hybrid algorithm (DPSO-BP) over other more standard algorithms in neural network training. The algorithms are compared using two different datasets, and the results are simulated.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1994.06a
• /
• pp.997-1002
• /
• 1994

Predictive neural network models are powerful speech recognition models based on a nonlinear pattern prediction. Those models can effectively normalize the temporal and spatial variability of speech signals. But those models suffer from poor discrimination between acoustically similar words. In this paper, we propose a discriminative training algorithm for predictive neural network models based on a generalized probabilistic descent (GPD) algorithm and minimum classification error formulation (MCEF). The Evaluation of our training algorithm on ten Korean digits shows its effectiveness by 40% reduction of recognition error.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.298-300
• /
• 2021

The Ministry of National Defense decided to build a realistic combat simulation training system based on virtual reality and augmented reality in accordance with the expansion of the scientific training system of "Defense Reform 2.0". The realistic combat simulation training system should be able to maximize the tension and training effect as in actual combat through engagement between trainers. In addition, it should be possible to increase the effectiveness of survival training at the same time as shooting training similar to actual combat through cover training. Previous studies are suitable techniques to improve the shooting precision of the trainee, but it is difficult to practice bilateral engagement like in actual combat, and it is particularly insufficient for combat shooting training using cover. Therefore, in this paper, we propose a S/W algorithm for generating a virtual avatar by recognizing the shooting posture of the opponent on the screen of the virtual shooting trainer. This S/W algorithm can recognize the trainer and the cover based on the depth information acquired through the depth sensor and estimate the trainer's posture.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.3
• /
• pp.654-661
• /
• 2011

This paper is about a personal identification algorithm using an ECG that has been studied by a few researchers recently. Previously published algorithm can be classified as two methods. One is the method that analyzes of ECG features and the other is the morphological analysis of ECG. The main characteristic of proposed algorithm can be classified the method of analysis ECG features. Proposed algorithm adopts DSTW(Down Slope Trace Wave) for extracting ECG features, and applies SVM(Support Vector Machine) to training and testing as a classifier algorithm. We choose 18 ECG files from MIT-BIH Normal Sinus Rhythm Database for estimating of algorithm performance. The algorithm extracts 100 heartbeats from each ECG file, and use 40 heartbeats for training and 60 heartbeats for testing. The proposed algorithm shows clearly superior performance in all ECG data, amounting to 93.89% heartbeat recognition rate and 100% ECG recognition rate.

• Journal of the Korea Computer Industry Society
• /
• v.2 no.4
• /
• pp.537-546
• /
• 2001

Griffth's and LCCMA which are Single-user detection adaptive algorithm are proposed for mitigate MAI(multiple access interference) and the near-far problem in direct-sequence spread-spectrum CDMA system and MOE Algorithm is proposed for MMSE(Minimum Mean-Square Error). This paper pertains to three types of Blind adaptive algorithms which can upgrade system functionality without the requirements from training sequence. It goes further to compare and analyze the functionalities of the algorithms as per number of interfering users or data update rate of the users. The simulation results was that LCCMA algorithm was superior to other algorithms in both areas. Blind application enabled a more flexible network design by eliminating the necessity of training sequence.

• ETRI Journal
• /
• v.34 no.5
• /
• pp.696-705
• /
• 2012

Accurate synthesis models based on artificial neural networks (ANNs) are proposed to directly obtain the physical dimensions of an asymmetric coplanar waveguide with conductor backing and substrate overlaying (ACPWCBSO). First, the ACPWCBSO is analyzed with the conformal mapping technique (CMT) to obtain the training data. Then, a modified genetic-algorithm-Levenberg-Marquardt (GA-LM) algorithm is adopted to train ANNs. In the algorithm, the maximal relative error (MRE) is used as the fitness function of the chromosomes to guarantee that the MRE is small, while the mean square error is used as the error function in LM training to ensure that the average relative error is small. The MRE of ANNs trained with the modified GA-LM algorithm is less than 8.1%, which is smaller than those trained with the existing GA-LM algorithm and the LM algorithm (greater than 15%). Lastly, the ANN synthesis models are validated by the CMT analysis, electromagnetic simulation, and measurements.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.12 no.10
• /
• pp.799-807
• /
• 2002

The vibration signal can give an indication of the condition of rotating machinery, highlighting potential faults such as unbalance, misalignment and bearing defects. The features in the vibration signal provide an important source of information for the faults diagnosis of rotating machinery. When additional training data become available after the initial training is completed, the conventional neural networks (NNs) must be retrained by applying total data including additional training data. This paper proposes the fault diagnostics algorithm using the ART-Kohonen network which does not destroy the initial training and can adapt additional training data that is suitable for the classification of machine condition. The results of the experiments confirm that the proposed algorithm performs better than other NNs as the self-organizing feature maps (SOFM) , learning vector quantization (LYQ) and radial basis function (RBF) NNs with respect to classification quality. The classification success rate for the ART-Kohonen network was 94 o/o and for the SOFM, LYQ and RBF network were 93 %, 93 % and 89 % respectively.

• Journal of Electrical Engineering and Technology
• /
• v.9 no.1
• /
• pp.52-59
• /
• 2014

Operator training simulator, within a training environment designed to understand the principles and behavior of the railway system with respect to operator's entries and predefined scenario, can provide a very strong benefit in facilitating operators' handling undesired operations. This simulator consists of computer system and applications, and the purpose of applications is to generate the power and voltage and analyze the AC substation and DC railway, respectively. This paper describes a novel approach to the new techniques for AC/DC hybrid simulation for the operator training simulator in the railway system. We first propose the structure the database of railway system. Then, topology processing and power flow using a linked-list method based on the proposed database, full or decoupled newton-rapshon methods are presented. Finally, the interface between the analysis for AC substation using a newton-rapshon method and the analysis for DC railway system using a time-interval power flow method is described. We have verified and tested the developed algorithm through the extensive testing for the proposed test system. To demonstrate the validity of the developed algorithm, comparative simulations between the proposed algorithm and PSS/E for the test system were conducted.