• The Bulletin of The Korean Astronomical Society
• /
• v.41 no.2
• /
• pp.54.2-54.2
• /
• 2016

We present a machine learning approach to estimating stellar atmospheric parameters, effective temperature (Teff), surface gravity (log g), and metallicity ([Fe/H]) for stars observed during the course of the Sloan Digital Sky Survey (SDSS). For training a neural network, we randomly sampled the SDSS data with stellar parameters available from SEGUE Stellar Parameter Pipeline (SSPP) to cover the parameter space as wide as possible. We selected stars that are not included in the training sample as validation sample to determine the accuracy and precision of each parameter. We also divided the training and validation samples into four groups that cover signal-to-noise ratio (S/N) of 10-20, 20-30, 30-50, and over 50 to assess the effect of S/N on the parameter estimation. We find from the comparison of the network-driven parameters with the SSPP ones the range of the uncertainties of 73~123 K in Teff, 0.18~0.42 dex in log g, and 0.12~0.25 dex in [Fe/H], respectively, depending on the S/N range adopted. We conclude that these precisions are high enough to study the chemical and kinematic properties of the Galactic disk and halo stars, and we will attempt to apply this technique to Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST), which plans to obtain about 8 million stellar spectra, in order to estimate stellar parameters.

• Journal of Communications and Networks
• /
• v.7 no.3
• /
• pp.243-247
• /
• 2005

We propose an efficient block least-mean-square (BLMS) adaptive algorithm, in conjunction with error control coding, for direct-sequence code division multiple access (DS-CDMA) systems. The proposed adaptive receiver incorporates decision feedback detection and channel encoding in order to improve the performance of the standard LMS algorithm in convolutionally coded systems. The BLMS algorithm involves two modes of operation: (i) The training mode where an uncoded training sequence is used for initial filter tap-weights adaptation, and (ii) the decision-directed where the filter weights are adapted, using the BLMS algorithm, after decoding/encoding operation. It is shown that the proposed adaptive receiver structure is able to compensate for the signal-to­noise ratio (SNR) loss incurred due to the switching from uncoded training mode to coded decision-directed mode. Our results show that by using the proposed adaptive receiver (with decision feed­back block adaptation) one can achieve a much better performance than both the coded LMS with no decision feedback employed. The convergence behavior of the proposed BLMS receiver is simulated and compared to the standard LMS with and without channel coding. We also examine the steady-state bit-error rate (BER) performance of the proposed adaptive BLMS and standard LMS, both with convolutional coding, where we show that the former is more superior than the latter especially at large SNRs ($SNR\;\geq\;9\;dB$).

• International journal of advanced smart convergence
• /
• v.9 no.2
• /
• pp.20-27
• /
• 2020

Sound event detection in real-world environments suffers from the interference of non-stationary and time-varying noise. This paper presents an adaptive noise reduction method for sound event detection based on non-negative matrix factorization (NMF). In this paper, we proposed a deep learning model that integrates Convolution Neural Network (CNN) with Non-Negative Matrix Factorization (NMF). To improve the separation quality of the NMF, it includes noise update technique that learns and adapts the characteristics of the current noise in real time. The noise update technique analyzes the sparsity and activity of the noise bias at the present time and decides the update training based on the noise candidate group obtained every frame in the previous noise reduction stage. Noise bias ranks selected as candidates for update training are updated in real time with discrimination NMF training. This NMF was applied to CNN and Hidden Markov Model(HMM) to achieve improvement for performance of sound event detection. Since CNN has a more obvious performance improvement effect, it can be widely used in sound source based CNN algorithm.

• Journal of Communications and Networks
• /
• v.15 no.4
• /
• pp.352-361
• /
• 2013

Combining multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) with a massive number of transmit antennas (massive MIMO-OFDM) is an attractive way of increasing the spectrum efficiency or reducing the transmission energy per bit. The effectiveness of Massive MIMO-OFDM is strongly affected by the channel state information (CSI) estimation method used. The overheads of training frame transmission and CSI feedback decrease multiple access channel (MAC) efficiency and increase the CSI estimation cost at a user station (STA). This paper proposes a CSI estimation scheme that reduces the training frame length by using a novel pilot design and a novel unitary matrix feedback method. The proposed pilot design and unitary matrix feedback enable the access point (AP) to estimate the CSI of the signal space of all transmit antennas using a small number of training frames. Simulations in an IEEE 802.11n channel verify the attractive transmission performance of the proposed methods.

• Journal of Ocean Engineering and Technology
• /
• v.18 no.6 s.61
• /
• pp.44-50
• /
• 2004

A small, highly-trained crew working on the ship's automation has contributed to the improvement of operations and the labor environment on board ship. However, at the same time, having a small crew adds more responsibility to the ship's officers to safely operate and manage the ship. With the use of information and computer technology, efforts are being made towards the development of a system that will concentrate important information from the various pieces of navigational equipment. The purpose of this study is to set up and implement an integrated marine data networking and communication system on the training-research ship. Information relating to navigation, engine and office automation are investigated and analyzed, and implementation methods associated with navigation, engine and management information system were designed and presented. In addition, the networking system of the navigational signal interface unit for the integrated communication system, and the data communication method between the ship and land are also discussed.

• Nuclear Engineering and Technology
• /
• v.51 no.7
• /
• pp.1784-1790
• /
• 2019

Accurate and consistent characterization of defects in steam generator tubes (SGT) in nuclear power plants is one of the key issues in the field of nondestructive testing since the large number of signals to be analyzed in a time-limited in-service inspection causes a serious problem in practice. This paper presents an effective approach to this difficult task of automated classification of motorized rotating pancake coil (MRPC) eddy current flaw acquired from tube specimens with deliberated defects using deep neural networks (DNN). This approach consists of five steps, namely, the data acquisition using the MRPC probe in the tube, the signal preprocessing to make data more suitable for training DNN, the data augmentation for boosting a training performance, the training of DNN, and finally demonstration of the trained DNN for discriminating the axial and circumferential defects. The high performance obtained in this study shows that DNN is useful for classification of defects in tubes from the MRPC eddy current signals even though the number of signals is very large.

• Communications for Statistical Applications and Methods
• /
• v.30 no.5
• /
• pp.501-516
• /
• 2023

Knowledge distillation (KD) refers to extracting knowledge from a large and complex model (teacher) and transferring it to a relatively small model (student). This can be done by training the teacher model to obtain the activation function values of the hidden or the output layers and then retraining the student model using the same training data with the obtained values. Recently, relational KD (RKD) has been proposed to extract knowledge about relative differences in training data. This method improved the performance of the student model compared to conventional KDs. In this paper, we propose a new method for RKD by introducing a new loss function for RKD. The proposed loss function is defined using the area difference between the teacher model and the student model in a specific hidden layer, and it is shown that the model can be successfully compressed, and the generalization performance of the model can be improved. We demonstrate that the accuracy of the model applying the method proposed in the study of model compression of audio data is up to 1.8% higher than that of the existing method. For the study of model generalization, we demonstrate that the model has up to 0.5% better performance in accuracy when introducing the RKD method to self-KD using image data.

• Nuclear Engineering and Technology
• /
• v.56 no.6
• /
• pp.2266-2273
• /
• 2024

A machine learning platform is proposed for the diagnosis of a severe accident progression in a nuclear power plant. To predict the key parameters for accident management including lost signals, a long short term memory (LSTM) network is proposed, where multiple accident scenarios are used for training. Training and test data were produced by MELCOR simulation of the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident at unit 3. Feature variables were selected among plant parameters, where the importance ranking was determined by a recursive feature elimination technique using RandomForestRegressor. To answer the question of whether a reduced order ML model could predict the complex transient response, we performed a systematic sensitivity study for the choices of target variables, the combination of training and test data, the number of feature variables, and the number of neurons to evaluate the performance of the proposed ML platform. The number of sensitivity cases was chosen to guarantee a 95 % tolerance limit with a 95 % confidence level based on Wilks' formula to quantify the uncertainty of predictions. The results of investigations indicate that the proposed ML platform consistently predicts the target variable. The median and mean predictions were close to the true value.

• Proceedings of the IEEK Conference
• /
• 2001.09a
• /
• pp.869-872
• /
• 2001

Blind identification and equalization of communication channel is important because it does not need training sequence, nor does it require a priori channel information. So, we can increase the bandwidth efficiency. The linear prediction error method is perhaps the most attractive in practice due to the insensitive to blind channel estimator and equalizer length mismatch as well as for its simple adaptive algorithms. In this paper, we propose method for fractionally spaced blind equalizer with arbitrary delay using one-step forward prediction error filter from second-order statistics of the received signals for SIMO channel. Our algorithm utilizes the forward prediction error as training sequences for data estimation and desired signal for channel estimation.

• Journal of Institute of Control, Robotics and Systems
• /
• v.2 no.1
• /
• pp.21-25
• /
• 1996

In this paper we suggest a general purpose RNN training algorithm which is derived on the optimal control concepts and variational methods. First, learning is regared as an optimal control problem, then using the variational methods we obtain optimal weights which are given by a two-point boundary-value problem. Finally, the modified gradient descent algorithm is applied to RNN for on-line training. This algorithm is intended to be used on learning complex dynamic mappings between time varing I/O data. It is useful for nonlinear control, identification, and signal processing application of RNN because its storage requirement is not high and on-line learning is possible. Simulation results for a nonlinear plant identification are illustrated.