• International Journal of Control, Automation, and Systems
• /
• 제3권2호
• /
• pp.152-158
• /
• 2005

Commercial lower limb prostheses or orthotics help patients achieve a normal life. However, patients who use such aids need prolonged training to achieve a normal gait, and their fatigability increases. To improve patient comfort, this study proposed a method of predicting gait angle using neural networks and EMG signals. Experimental results using our method show that the absolute average error of the estimated gait angles is $0.25^{\circ}$. This performance data used reference input from a controller for the lower limb orthotic or prosthesis controllers while the patients were walking.

• Proceedings of the KIEE Conference
• /
• 대한전기학회 2008년도 심포지엄 논문집 정보 및 제어부문
• /
• pp.213-214
• /
• 2008

This paper considers design of a robust controller that alleviates disturbance effects and compensates performance degradation of plants with time-delay. Disturbance observer(DOB) approach as a tool of robust control has been widely employed in industry. However, since the time-delay makes the plant non-minimum phase, classical DOB cannot be applied directly to the time-delay system. By using a new DOB structure for non-minimum phase systems together with the Smith Predictor, we propose a new control algorithm for reducing the effects of disturbance and time-delay of the system.

• English Language & Literature Teaching
• /
• 제13권1호
• /
• pp.27-46
• /
• 2007

Teachers can maximize students' L2 learning by knowing preferred learning styles. This paper presents the results of a survey that asked 309 English learners to identify their perceptual learning style preferences. It further compared students' favored learning styles in terms of their gender and major field of study and explored a possible link between learning styles and English achievement. Collected data using Reid's (1995) questionnaire were analyzed by descriptive statistics, MANOVA, ANOVA, correlations, multiple regressions including squared partial correlations, and Cronbach's alpha. The results indicated that Korean students favored English learning in group regardless of gender, while their preferred mode of learning was significantly different in regard to their major field of study. Certain learning styles might be profitable for English achievement. Multiple regression analyses revealed that individual mode of learning was the best predictor of students' English achievement. It furthermore showed significant relationships between visual and individual styles of learning and English performance. The findings of the study reflected students' English learning context in which English native-speaking teachers frequently used communicative pair and small group activities for speaking practices that were consonant with students' learning styles.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2004년도 ICCAS
• /
• pp.395-400
• /
• 2004

In most physical processes, the transfer function includes time-delay, and in the general distributed control system using computer network, there exists inherent time-delay caused by the spatial separation between controllers and actuators. This work deals with the synthesis of a discrete-time controller for a nonlinear system and proposes a new effective method to compensate the influence of input time-delay. The controller is synthesized by using input/output linearization. Under the circumstance that input time-delay exists, the system response has more overshoot and tends to diverge. For these reasons, the controller has to produce future input value that will be needed for the system. In order to calculate the future input value, some predictors are adopted. Using the discretization via Euler's method, numerical simulations about the Van der Pol system are performed to evaluate the performance of the proposed method.

• Proceedings of the IEEK Conference
• /
• 대한전자공학회 2000년도 추계종합학술대회 논문집(4)
• /
• pp.163-166
• /
• 2000

In this paper, we will present a lossy data compression method for coding multispectral images. The proposed method uses both spatial and spectra] correlation inherent in multispectral images. First, band 2 and band 6 are vector quantized. Secondly, band 4 is estimated with the quantized band 2 using the predictive coding. Errors of band 4 are encoded at a second stage based on the magnitude of the errors. Thirdly, remaining bands are calculated with the quantized band 2 and band 4. Errors of residual bands are wavelet transformed and then we apply the SPIHT coding on the transformed coefficients. We classify classes without extra information transmitting and then use linear predictor. And errors can be encoded by SPIHT coding at any target rate we are want. It is shown that this method has better performance than FPVQ. Average PSNR rises 0.645 dB at the same bit rate.

• Journal of Institute of Control, Robotics and Systems
• /
• 제6권6호
• /
• pp.441-446
• /
• 2000

This paper presents an approach for the prediction of optimal heating load using a diagonal recurrent neural networks(DRNN) and data base system of outdoor temperature. In the DRNN, a dynamic backpropagation(DBP) with delta-bar-delta teaming method is used to train an optimal heating load identifier. And the data base system is utilized for outdoor temperature prediction. Compared to other kinds of methods, the proposed method gives better prediction performance of heating load. Also a hardware for the controller is developed using a microprocessor. The experimental results show that prediction enhancement for heating load can be achieved with the proposed method regardless of the its inherent nonlinearity and large time constant.

• Korean Management Science Review
• /
• 제11권2호
• /
• pp.133-146
• /
• 1994

This study compares the performance of neural networks and ordinary least squares regression with quality-control processes. We examine the applicability of neural networks because they do not require any assumptions regarding either the functional from of the underlying process or the distribution of errors. The coefficient of determination($R^2$), mean absolute deviation(MAD), and the mean squared error(MSE) metrics indicate that neural networks are a viable and can be a superior technique. We also demonstrate that an assessment of the magnitude of the neural notwork input layer cumulative weights can be used to determine the relative importance of predictor variables.

• Proceedings of the KIPE Conference
• /
• 전력전자학회 2007년도 추계학술대회 논문집
• /
• pp.163-165
• /
• 2007

Direct Instantaneous Pressure Control(DIPC) method of SRM using pressure predict method is presented in this paper. A hydraulic pump system has an inherent defect that its dynamic behavior causes by interaction between the sensor and hydraulic load. It will make sometimes the whole system become oscillatory and unstable. Proposed system integrates direct instantaneous torque control (DITC) and Smith predictor to improve dynamic performance and stabilization. The proposed hydraulic oil pump system is verified by computer simulation and experimental results.

• International Journal of Contents
• /
• 제9권4호
• /
• pp.11-15
• /
• 2013

"Protein Folding Problem" is considered to be one of the "Great Challenges of Computer Science" and prediction of disordered protein is an important part of the protein folding problem. Machine learning models can predict the disordered structure of protein based on its characteristic of "learning from examples". Among many machine learning models, we investigate the possibility of multilayer perceptron (MLP) as the predictor of protein disorder. The investigation includes a single hidden layer MLP, multi hidden layer MLP and the hierarchical structure of MLP. Also, the target node cost function which deals with imbalanced data is used as training criteria of MLPs. Based on the investigation results, we insist that MLP should have deep architectures for performance improvement of protein disorder prediction.

• Industrial Engineering and Management Systems
• /
• 제13권4호
• /
• pp.421-431
• /
• 2014

The market basket data in the form of a binary user-item matrix or a binary item-user matrix can be modelled as a binary classification problem. The binary logistic regression approach tackles the binary classification problem, where principal components are predictor variables. If users or items are sparse in the training data, the binary classification problem can be considered as a cold-start problem. The binary logistic regression approach may not function appropriately if the principal components are inefficient for the cold-start problem. Assuming that the market basket data can also be considered as a special regression problem whose response is either 0 or 1, we propose three supervised learning approaches: random forest regression, random forest classification, and elastic net to tackle the cold-start problem, comparing the performance in a variety of experimental settings. The experimental results show that the proposed supervised learning approaches outperform the conventional approaches.