• Proceedings of the KIEE Conference
• /
• 2007.04a
• /
• pp.3-5
• /
• 2007

This paper presents decision feedback equalizers using a recurrent neural network trained algorithm using extended Kalman filter(EKF) and sigma-point Kalman filter(SPKF). EKF is propagated, analytically through the first-order linearization of the nonlinear system. This can introduce large errors in the true posterior mean and covariance of the Gaussian random variable. The SPKF addresses this problem by using a deterministic sampling approach. The features of the proposed recurrent neural equalizer And we investigate the bit error rate(BER) between EKF and SPKF.

• Communications for Statistical Applications and Methods
• /
• v.27 no.3
• /
• pp.313-326
• /
• 2020

Deep Learning is the most important key to the development of Artificial Intelligence (AI). There are several distinguishable architectures of neural networks such as MLP, CNN, and RNN. Among them, we try to understand one of the main architectures called Recurrent Neural Network (RNN) that differs from other networks in handling sequential data, including time series and texts. As one of the main tasks recently in Natural Language Processing (NLP), we consider Neural Machine Translation (NMT) using RNNs. We also summarize fundamental structures of the recurrent networks, and some topics of representing natural words to reasonable numeric vectors. We organize topics to understand estimation procedures from representing input source sequences to predict target translated sequences. In addition, we apply multiple translation models with Gated Recurrent Unites (GRUs) in Keras on English-Korean sentences that contain about 26,000 pairwise sequences in total from two different corpora, colloquialism and news. We verified some crucial factors that influence the quality of training. We found that loss decreases with more recurrent dimensions and using bidirectional RNN in the encoder when dealing with short sequences. We also computed BLEU scores which are the main measures of the translation performance, and compared them with the score from Google Translate using the same test sentences. We sum up some difficulties when training a proper translation model as well as dealing with Korean language. The use of Keras in Python for overall tasks from processing raw texts to evaluating the translation model also allows us to include some useful functions and vocabulary libraries as well.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2004.05b
• /
• pp.1285-1289
• /
• 2004

Elman Discrete Recurrent Neural Networks Model(EDRNNM) was used to be a suitable short-term hydrological forecasting tool yielding a very high degree of flood stage forecasting accuracy at Musung station of Wi-stream one of IHP representative basins in South Korea. A relative new approach method has recurrent feedback nodes and virtual small memory in the structure. EDRNNM was trained by using two algorithms, namely, LMBP and RBP The model parameters, optimal connection weights and biases, were estimated during training procedure. They were applied to evaluate model validation. Sensitivity analysis test was also performed to account for the uncertainty of input nodes information. The sensitivity analysis approach could suggest a reduction of one from five initially chosen input nodes. Because the uncertainty of input nodes information always result in uncertainty in model results, it can help to reduce the uncertainty of EDRNNM application and management in small catchment.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.1439-1442
• /
• 2010

Time-lag recurrent neural networks model (Time-Lag RNNM) is used to estimate daily pan evaporation (PE) using limited climatic variables such as max temperature ($T_{max}$), min temperature ($T_{min}$), mean wind speed ($W_{mean}$) and mean relative humidity ($RH_{mean}$). And, for the performances of Time-Lag RNNM, it is composed of training and test performances, respectively. The training and test performances are carried out using daily time series data, respectively. From this research, we evaluate the impact of Time-Lag RNNM for the modeling of the nonlinear time series data. We should, thus, construct the credible data of the daily PE using Time-Lag RNNM, and can suggest the methodology for the irrigation and drainage networks system. Furthermore, this research represents that the strong nonlinear relationship such as pan evaporation modeling can be generalized using Time-Lag RNNM.

• Journal of Ocean Engineering and Technology
• /
• v.35 no.4
• /
• pp.287-295
• /
• 2021

This paper describes a recurrent neural network (RNN) for the fault classification of a blade pitch system of a spar-type floating wind turbine. An artificial neural network (ANN) can effectively recognize multiple faults of a system and build a training model with training data for decision-making. The ANN comprises an encoder and a decoder. The encoder uses a gated recurrent unit, which is a recurrent neural network, for dimensionality reduction of the input data. The decoder uses a multilayer perceptron (MLP) for diagnosis decision-making. To create data, we use a wind turbine simulator that enables fully coupled nonlinear time-domain numerical simulations of offshore wind turbines considering six fault types including biases and fixed outputs in pitch sensors and excessive friction, slit lock, incorrect voltage, and short circuits in actuators. The input data are time-series data collected by two sensors and two control inputs under the condition that of one fault of the six types occurs. A gated recurrent unit (GRU) that is one of the RNNs classifies the suggested faults of the blade pitch system. The performance of fault classification based on the gate recurrent unit is evaluated by a test procedure, and the results indicate that the proposed scheme works effectively. The proposed ANN shows a 1.4% improvement in its performance compared to an MLP-based approach.

• Proceedings of the IEEK Conference
• /
• 2008.06a
• /
• pp.667-668
• /
• 2008

This paper demonstrates that the recurrent neural networks can be used successfully for the identification of finite automata (FAs). A new type of recurrent neural network (RNN) is proposed and the offline training algorithm, regulated Levenberg-Marquadt (LM) algorithm, for the network is developed. Simulation result shows that the identification and the extraction of FAs are practically achievable.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.5
• /
• pp.552-559
• /
• 2005

Recurrent neural networks(RNNs) trained with gradient based such as real time recurrent learning(RTRL) has a drawback of slor convergence rate. This algorithm also needs the derivative calculation which is not trivialized in error back propagation process. In this paper a derivative free Kalman filter, so called the unscented Kalman filter(UKF), for training a fully connected RNN is presented in a state space formulation of the system. A derivative free Kalman filler learning algorithm makes the RNN have fast convergence speed and good tracking performance without the derivative computation. Through experiments of nonlinear channel equalization, performance of the RNNs with a derivative free Kalman filter teaming algorithm is evaluated.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.24
• /
• pp.10997-10998
• /
• 2015

• Journal of Korean Ophthalmic Optics Society
• /
• v.16 no.3
• /
• pp.345-350
• /
• 2011

Purpose: This was a study of the effects of the vision training on binocular function and sensory anomaly for intermittent exotropia and recurrent intermittent exotropia patients. Mehtods: The change was examined weekly at the clinic during the 8 week vision training. The sample group of 17 patients, intermittent exotropia and recurrent intermittent exotropia patients without amblyopia, were chosen from the patients with the short and the long-distance deviation angle p=0.558 and p=0.338, and also greater than 10${\Delta}$, respectively. Results: The longdistance deviation angle was reduced by 67.9(${\pm}$32.7)%, and the short-distance deviation angle by 75.4(${\pm}$32.9)% after the vision training. There is an improvement on the binocular function and sensory anomaly according to the fusion reverse and Worth 4 dot tests. Conclusions: This study shows that the vision training is effective for phoria, function and sensory anomalyon of the intermittent extropia and recurrent intermittent exotropia patients.

• The Journal of the Acoustical Society of Korea
• /
• v.42 no.5
• /
• pp.395-401
• /
• 2023

Sound Event Detection (SED) aims to identify not only sound category but also time interval for target sounds in an audio waveform. It is a critical technique in field of acoustic surveillance system and monitoring system. Recently, various models have introduced through Detection and Classification of Acoustic Scenes and Events (DCASE) Task 4. This paper explored how to design optimal parameters of DenseNet based model, which has led to outstanding performance in other recognition system. In experiment, DenseRNN as an SED model consists of DensNet-BC and bi-directional Gated Recurrent Units (GRU). This model is trained with Mean teacher model. With an event-based f-score, evaluation is performed depending on parameters, related to model architecture as well as model training, under the assessment protocol of DCASE task4. Experimental result shows that the performance goes up and has been saturated to near the best. Also, DenseRNN would be trained more effectively without dropout technique.