• The KIPS Transactions:PartD
• /
• v.8D no.3
• /
• pp.241-246
• /
• 2001

Area of software measurement in software engineering is active more than thirty years. There is a huge collection of researches but still no a concrete software cost estimation model. If we want to measure the cost-effort of a software project, we need to estimate the size of the software. A number of software metrics are identified in the literature ; the most frequently cited measures are LOC(line of code) and FPA(function point analysis). The FPA approach has features that overcome the major problems with using LOC as a measure of system size. This paper presents an neural networks(NN) models that related software development effort to software size measured in FPs and function element types. The research describes appropriate NN modeling in the context of a case study for 24 software development projects. Also, this paper compared the NN model with a regression analysis model and found the NN model has better estimative accuracy.

• Journal of Korea Society of Industrial Information Systems
• /
• v.25 no.2
• /
• pp.129-136
• /
• 2020

Recognizing finger movements have been used as a intuitive way of human-computer interaction. In this study, we implement an wearable device for finger motion recognition and evaluate the accuracy of several ML (Machine learning) techniques. Not only HMM (Hidden markov model) and DTW (Dynamic time warping) techniques that have been traditionally used as time series data analysis, but also NN (Neural network) technique are applied to compare and analyze the accuracy of each technique. In order to minimize the computational requirement, we also apply the pre-processing to each ML techniques. Our extensive evaluations demonstrate that the NN-based gesture recognition system achieves 99.1% recognition accuracy while the HMM and DTW achieve 96.6% and 95.9% recognition accuracy, respectively.

• Journal of Electrical Engineering and Technology
• /
• v.10 no.1
• /
• pp.408-421
• /
• 2015

Because the wheel of V-belt continuously variable transmission (CVT) system driven by permanent magnet synchronous motor (PMSM) has much unknown nonlinear and time-varying characteristics, the better control performance design for the linear control design is a time consuming job. In order to overcome difficulties for design of the linear controllers, a hybrid recurrent Chebyshev neural network (NN) control system is proposed to control for a PMSM servo-driven V-belt CVT system under the occurrence of the lumped nonlinear load disturbances. The hybrid recurrent Chebyshev NN control system consists of an inspector control, a recurrent Chebyshev NN control with adaptive law and a recouped control. Moreover, the online parameters tuning methodology of adaptive law in the recurrent Chebyshev NN can be derived according to the Lyapunov stability theorem and the gradient descent method. Furthermore, the optimal learning rate of the parameters based on discrete-type Lyapunov function is derived to achieve fast convergence. The recurrent Chebyshev NN with fast convergence has the online learning ability to respond to the system's nonlinear and time-varying behaviors. Finally, to show the effectiveness of the proposed control scheme, comparative studies are demonstrated by experimental results.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.5
• /
• pp.956-961
• /
• 2006

Keystroke time intervals can be a discriminating feature in the verification and identification of computer users. This paper presents a comparison result obtained using several classification methods including k-NN (k-Nearest Neighbor), back-propagation neural networks, and Bayesian classification for keystroke identification. Performance of k-NN classification was best with small data samples available per user, while Bayesian classification was the most superior to others with large data samples per user. Thus, for web-based on-line identification of users, it seems to be appropriate to selectively use either k-NN or Bayesian method according to the number of keystroke samples accumulated by each user.

• Journal of the Institute of Convergence Signal Processing
• /
• v.3 no.2
• /
• pp.30-36
• /
• 2002

This paper describes a nonlinear adaptive noise canceler(ANC) using neural networks(NN) based on filter to make up for the drawback of the conventional ANC with the linear adaptive filter. The proposed ANC was tested its noise rejection performance using broadband time-varying noise signal and compared with the ANC of TDL linear filter. Experimental results show that in cases of nonlinear correlations between the noise of primary input and reference input, the neural network based ANC outperforms the linear ANC with respect to mean square error It is also verified that the recurrent NN adaptive filter is superior to the feedforward NN filter. Thus, we identify that the NN adaptive filter is more effective than the linear adaptive filter for rejection of broadband time-varying noise in the ANC.

• Journal of Korea Water Resources Association
• /
• v.36 no.4
• /
• pp.633-645
• /
• 2003

The techniques of GIS and remote sensing are being applied to hydrology, geomorphology and various field of studies are performed by many researcher, related those techniques. In this paper, curve number change detection is tested according to soil map and land cover in mountain area. Neural networks method is applied for land cover classification and GIS for curve number calculation. The first, sample area are selected and tested land cover classification, NN(84.1％) is superior to MLC(80.9％). So we selected NN with land cover classifier. The second, curve number from the land cover by neural network classifier(57) is compared with that(curve number) from the land cover by manual work(55). Two values are so similar. The third, curve number classified by NN in sample area was applied and tested to whole study area. As results of this study, it is shown that curve number is more exact and efficient by using NN and GIS technique than by (using) manual work.

• Communications for Statistical Applications and Methods
• /
• v.26 no.6
• /
• pp.591-610
• /
• 2019

Deep Learning is one of the machine learning methods to find features from a huge data using non-linear transformation. It is now commonly used for supervised learning in many fields. In particular, Convolutional Neural Network (CNN) is the best technique for the image classification since 2012. For users who consider deep learning models for real-world applications, Keras is a popular API for neural networks written in Python and also can be used in R. We try examine the parameter estimation procedures of Deep Neural Network and structures of CNN models from basics to advanced techniques. We also try to figure out some crucial steps in CNN that can improve image classification performance in the CIFAR10 dataset using Keras. We found that several stacks of convolutional layers and batch normalization could improve prediction performance. We also compared image classification performances with other machine learning methods, including K-Nearest Neighbors (K-NN), Random Forest, and XGBoost, in both MNIST and CIFAR10 dataset.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.2
• /
• pp.293-300
• /
• 2011

A number of techniques have been developed for estimation of speed or position in motor drives. The accuracy of these techniques is affected by the variation of motor parameters such as the stator resistance, stator inductance or torque constant. This paper is proposed a neural network based estimator for torque and stator resistance and adaptive fuzzy learning contrroller(AFLC) for speed control in IPMSM Drives. AFLC is chaged fuzzy rule base by rule base modifier for robust control of IPMSM. The neural weights are initially chosen randomly and a model reference algorithm adjusts those weights to give the optimum estimations. The neural network estimator is able to track the varying parameters quite accurately at different speeds with consistent performance. The neural network parameter estimator has been applied to slot and flux linkage torque ripple minimization of the IPMSM. The validity of the proposed parameter estimator and AFLC is confirmed by comparing to conventional algorithm.

• Journal of the Korean Society for Precision Engineering
• /
• v.34 no.2
• /
• pp.125-131
• /
• 2017

2-Axis Pan and Tilt Motion Platform, a complex multivariate non-linear system, may incur any disturbance, thus requiring system controller with robustness against various disturbances. In this study, we designed an adaptive backstepping compensated controller by estimating the disturbance and error using the Radial Basis Function Neural Network (RBF NN). In this process, Uniformly Ultimately Bounded (UUB) was demonstrated via Lyapunov and stability was confirmed. By generating progressive disturbance to the irregular frequency and amplitude changes, it was verified for various environmental disturbances. In addition, by setting the RBF NN input vector to the minimum, the estimated disturbance compensation process was analyzed. Only two input vectors facilitated compensatory function of RBF NN via estimating the modeling and control error values as well as irregular disturbance; the application of the process resulted in improved backstepping controller performance that was confirmed through simulation.

• Proceedings of the Korean Vacuum Society Conference
• /
• 2010.02a
• /
• pp.262-262
• /
• 2010

반도체 제조공정 중 식각공정(Etching)은 웨이퍼표면으로부터 화학적, 물리적으로 불필요한 물질들을 선택적으로 제거하는 방법이다. 식각공정 중 하나인 플라즈마 식각(Plasma etching) 공정에서 오버식각(over-etching) 과언더식각(under-etching) 되는것을피하기위해서통계적인방법을기준으로식각종료점(endpoint)를 결정한다. 본 논문의 목표는 통계적인 분석방법을 이용하지 않고 실시간 식각 데이터(realtime etching data)를 사용해서 식각 종료점을 검출하는 것이다. 식각 데이터는 시계열 데이터(time-series data)이기 때문에 간단한 구조와 적은 계산량으로 빠른 수렴속도와 좋은 안정도를 가진 Radial Basis Function Neural Network's (RBF-NN) 를 이용하여 시계열 모델(time-series model)을 구현 하였다. 광학방사분광기(Optical Emission Spectroscopy: OES)로부터 나온 6개의 데이터 세트중에서 4개의 데이터 세트는 RBF-NN을 학습하는데 사용되고 2개의 데이터 세트는 모델의 성과를 시험해 보기 위하여 사용하였다. 학습을 위한 데이터들은 Matrix화 시켜서 목표값을 설정하여 학습시킨다. 실험한 결과 학습한 RBF-NN 모형이 식각 종료점(endpoint)를 정확하게 검출된다는 것을 보여준다.