• The Journal of the Acoustical Society of Korea
• /
• v.33 no.4
• /
• pp.268-272
• /
• 2014

This paper presents a method for speech recognition using multi-section vector-quantization (MSVQ) and time-delay recurrent neural network (TDTNN). The MSVQ generates the codebook with normalized uniform sections of voice signal, and the TDRNN performs the speech recognition using the MSVQ codebook. The TDRNN is a time-delay recurrent neural network classifier with two different representations of dynamic context: the time-delayed input nodes represent local dynamic context, while the recursive nodes are able to represent long-term dynamic context of voice signal. The cepstral PLP coefficients were used as speech features. In the speech recognition experiments, the MSVQ/TDRNN speech recognizer shows 97.9 % word recognition rate for speaker independent recognition.

• Journal of KIISE:Computer Systems and Theory
• /
• v.28 no.9
• /
• pp.461-468
• /
• 2001

Though shared virtual memory (SVM) system promise low cost solutions for high performance computing they suffer from long memory latencies. These latencies are usually caused by repetitive invalidations on shared data. Since shared data are accessed through synchronization and the patterns by which threads synchronizes are repetitive, a prefetching scheme bases on such repetitiveness would reduce memory latencies. Based on this observation, we propose a prefetching technique which predicts future access behavior by analyzing access history per synchronization variable. Our technique was evaluated on an 8-node SVM system using the SPLASH-2 benchmark. The results show the our technique could achieve 34%~45% reduction in memory access latencies.

• Proceedings of the Korean Institute of Landscape Architecture Conference
• /
• 2022.10a
• /
• pp.106-107
• /
• 2022

• Journal of Navigation and Port Research
• /
• v.35 no.1
• /
• pp.83-91
• /
• 2011

The accuracy of forecasting is remarkably important to reduce total cost or to increase customer services, so it has been studied by many researchers. In this paper, the artificial neural network (ANN), one of the most popular nonlinear forecasting methods, is compared with autoregressive integrated moving average(ARIMA) model through performing a prediction of container traffic. It uses a hybrid methodology that combines both the linear ARIAM and the nonlinear ANN model to improve forecasting performance. Also, it compares the methodology with other models in performance for prediction. In designing network structure, this work specially applies the genetic algorithm which is known as the effectively optimal algorithm in the huge and complex sample space. It includes the time delayed neural network (TDNN) as well as multi-layer perceptron (MLP) which is the most popular neural network model. Experimental results indicate that both ANN and Hybrid models outperform ARIMA model.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.40 no.3
• /
• pp.303-314
• /
• 2020

The hydrological characteristics of watersheds and hydraulic systems of urban and river floods are highly nonlinear and contain uncertain variables. Therefore, the predicted time series of rainfall-runoff data in flood analysis is not suitable for existing neural networks. To overcome the challenge of prediction, a NARX (Nonlinear Autoregressive Exogenous Model), which is a kind of recurrent dynamic neural network that maximizes the learning ability of a neural network, was applied to forecast a flood in real-time. At the same time, NARX has the characteristics of a time-delay neural network. In this study, a hydrological model was constructed for the Taehwa river basin, and the NARX time-delay parameter was adjusted 10 to 120 minutes. As a result, we found that precise prediction is possible as the time-delay parameter was increased by confirming that the NSE increased from 0.530 to 0.988 and the RMSE decreased from 379.9 ㎥/s to 16.1 ㎥/s. The machine learning technique with NARX will contribute to the accurate prediction of flow rate with an unexpected extreme flood condition.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.20 no.5
• /
• pp.1013-1019
• /
• 2016

This paper describes a novel Context-awareness Markov Chain Prediction (CMCP) algorithm based on movement prediction using Markov chain in Delay Tolerant Network (DTN). The existing prediction models require additional information such as a node's schedule and delivery predictability. However, network reliability is lowered when additional information is unknown. To solve this problem, we propose a CMCP model based on node behaviour movement that can predict the mobility without requiring additional information such as a node's schedule or connectivity between nodes in periodic interval node behavior. The main contribution of this paper is the definition of approximate speed and direction for prediction scheme. The prediction of node movement forwarding path is made by manipulating the transition probability matrix based on Markov chain models including buffer availability and given interval time. We present simulation results indicating that such a scheme can be beneficial effects that increased the delivery ratio and decreased the transmission delay time of predicting movement path of the node in DTN.

• The Korean Journal of Nuclear Medicine
• /
• v.32 no.3
• /
• pp.259-265
• /
• 1998

Purpose: We studied early rest/24 hour delay T1-201 perfusion SPECT for prediction of wall motion improvement after reperfusion in patients with acute myocardial infarction. Materials and Methods: Among 17 patients (male/female= 11/6, age: $59{\pm}13$) with acute myocardial infarction, 15 patients were treated with percutaneous transcoronary angioplasty (direct:2, delay: 11) and intravenous urokinase (2). Spontaneous resolution occurred in infarct-related arteries of 2 patients. We confirmed TIMI 3 flow of infarct-related artery after reperfusion in all patients with coronary angiography. We performed rest T1-201 perfusion SPECT less then 6 hours after reperfusion and delay T1-201 perfusion SPECT next day. T1-201 uptake was visually graded as 4 point score from normal (0) to severe defect (3). Rest T1-201 uptake ${\le}2$ or combination of rest T1-201 uptake ${\le}2$ or late reversibility were considered to be viable. Myocardial wall motion was graded as 5 point score from normal (1) to dyskinesia (5). Myocardial wall motion was considered to be improved when a segment showed an improvement ${\ge} 1$ grade in follow up echo compared with the baseline values. Results: Among 98 segments with wall motion abnormality, the severity of myocardial wall motion decrease was as follow: mild hypokinesia: 18/98 (18%), severe hypokinesia: 28/98 (29%), akinesia: 51/98 (52%), dyskinesia: 1/98 (1%). The wall motion improved in 85%. Redistribution (13%), and reverse redistribution (4%) were observed in 24 hour delay SPECT. Positive predictive value (PPV) and negative predictive value (NPV) of combination of late reversibility and rest T1-201 uptake were 99%, and 54%. PPV and NPV of rest T1-201 uptake were 100% and 52% respectively. Predictive values of combination of rest T1-201 uptake and late reversibility were not significantly different compared with predictive values of rest T1-201 uptake only. Conclusion: We conclude that early T1-201 perfusion SPECT predict myocardial wall motion improvement with excellent positive but relatively low negative predictive values in patients with acute myocardial infarction after reperfusion.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.11
• /
• pp.755-761
• /
• 2017

In railways powered by AC power, the main circuit breaker (MCB) is used for supplying the electric power to the catenary of the vehicle. Generally, the main circuit breaker is located between the pantograph and the main transformer, and the phase of the power applied to the vehicle changes according to the operation timing of the main circuit breaker. The operation of the main circuit breaker should be actively controlled according to the phase of the power source, since the phase of the power causes unintended transient states in the vehicle's electrical system in the form of an inrush current and surge voltage. However, the MCB has a delay time when it operates which is not constant. Therefore, an intelligent controller is needed to predict the operation delay time and control the opening and closing of the MCB.

• Convergence Security Journal
• /
• v.7 no.2
• /
• pp.17-25
• /
• 2007

In this paper, an algorithm that provisions proportional differentiation of packet delays is proposed with an objective for enhancing quality of service (QoS) in future packet networks. It features an adaptive scheme that adjusts the target delay every time slot to compensate the deviation from the target delay which is caused by the prediction error on the traffic to be arrived in the next time slot. It predicts the traffic to be arrived at the beginning of a time slot and measures the actual arrived traffic at the end of the time slot. The difference between them is utilized to the delay control operation for the next time slot to offset it. As it compensates the prediction error continuously, it shows superior adaptability to the bursty traffic as well as the exponential rate traffic. It is demonstrated through simulations that the algorithm meets the quantitative delay bounds and shows superiority to the traffic fluctuation in comparison with the conventional non-adaptive mechanism. The algorithm is implemented with VHDL on a Xilinx Spartan XC3S1500 FPGA and the performance is verified under the test board based on the XPC860P CPU.

• The Proceedings of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.5 no.4
• /
• pp.52-59
• /
• 1991

본 논문은 유한시간 정정응답 제어이론을 이용한 직류전동기의 속도제어에 대하여 논하였다. 유한시간 정정응답 제어계는 이산시간제어를 적용하므로 제어량 포화현상과 검출지연 문제가 발생하여 계통의 불안정을 초래한다. 이러한 문제를 해결하기 위하여 포화상태에서도 고속응답이 가능하도록 예측제어를 적용한 보상기를 제안한다. 실험 결과 지령치가 포화되지 않은 상태에서는 1샘플링시간으로 정정할 수 있었다. 지령치가 포화한 상태에서는 포화로부터 벗어난 후 1샘플링시간으로 정정할 수 있었다. 또한 예측제어를 적용하므로써 과도시의 오우버슈우트가 억제된 고속정정이 가능함을 알 수 있었다.