• Proceedings of the Korean Powder Metallurgy Institute Conference
• /
• 2006.09a
• /
• pp.462-463
• /
• 2006

Monodispersed and nano-sized Ni powders were synthesized from aqueous nickel sulfate hexahydrate $(NiSO_4{\cdot}6H_2O)$ inside nonionic polymer network by using wet chemical reduction process. The sucrose was used as a nonionic polymer network source. The effect of reaction conditions such as the amount of sucrose and a various reaction temperature, nickel sulfate hexahydrate molarity. The influence of a nonionic polymer network on the particle size of the prepared Ni powders was characterized by means of X-ray diffraction (XRD), scanning electron microscopy (SEM), and particle size analysis (PSA). The results showed that the obtained Ni powders were strong by dependent of the reaction conditions. In particular, the Ni powders prepared inside a nonionic polymer network had smooth spherical shape and narrow particle size distribution.

• Communications for Statistical Applications and Methods
• /
• v.31 no.2
• /
• pp.213-234
• /
• 2024

In recent decades, increasing research attention has been directed toward predicting the price of stocks in financial markets using deep learning methods. For instance, recurrent neural network (RNN) is known to be competitive for datasets with time-series data. Long short term memory (LSTM) further improves RNN by providing an alternative approach to the gradient loss problem. LSTM has its own advantage in predictive accuracy by retaining memory for a longer time. In this paper, we combine both supervised and unsupervised dimension reduction methods with LSTM to enhance the forecasting performance and refer to this as a dimension reduction based LSTM (DR-LSTM) approach. For a supervised dimension reduction method, we use methods such as sliced inverse regression (SIR), sparse SIR, and kernel SIR. Furthermore, principal component analysis (PCA), sparse PCA, and kernel PCA are used as unsupervised dimension reduction methods. Using datasets of real stock market index (S&P 500, STOXX Europe 600, and KOSPI), we present a comparative study on predictive accuracy between six DR-LSTM methods and time series modeling.

• Publications of The Korean Astronomical Society
• /
• v.28 no.1
• /
• pp.1-6
• /
• 2013

Real-time data reduction pipeline for the Korea Microlensing Telescope Network (KMTNet) was developed by Korea Astronomy and Space Science Institute (KASI). The main goal of the data reduction pipeline is to find variable objects and to record their light variation from the large amount of observation data of about 200 GB per night per site. To achieve the goal we adopt three strategic implementations: precision pointing of telescope using the cross correlation correction for target fields, realtime data transferring using kernel-level file handling and high speed network, and segment data processing architecture using the Sun-Grid engine. We tested performance of the pipeline using simulated data which represent the similar circumstance to CTIO (Cerro Tololo Inter-American Observatory), and we have found that it takes about eight hours for whole processing of one-night data. Therefore we conclude that the pipeline works without problem in real-time if the network speed is high enough, e.g., as high as in CTIO.

• Smart Structures and Systems
• /
• v.5 no.4
• /
• pp.415-426
• /
• 2009

This study focuses on the concept of multi-scale wireless sensor networks for damage detection in civil infrastructure systems by first over viewing the general network philosophy and attributes in the areas of data acquisition, data reduction, assessment and decision making. The data acquisition aspect includes a scalable wireless sensor network acquiring acceleration and strain data, triggered using a Restricted Input Network Activation scheme (RINAS) that extends network lifetime and reduces the size of the requisite undamaged reference pool. Major emphasis is given in this study to data reduction and assessment aspects that enable a decentralized approach operating within the hardware and power constraints of wireless sensor networks to avoid issues associated with packet loss, synchronization and latency. After over viewing various models for data reduction, the concept of a data-driven Bivariate Regressive Adaptive INdex (BRAIN) for damage detection is presented. Subsequent examples using experimental and simulated data verify two major hypotheses related to the BRAIN concept: (i) data-driven damage metrics are more robust and reliable than their counterparts and (ii) the use of heterogeneous sensing enhances overall detection capability of such data-driven damage metrics.

• Information and Communications Magazine
• /
• v.32 no.10
• /
• pp.31-40
• /
• 2015

This paper investigates several existing techniques to reduce high-availability seamless redundancy (HSR) traffic. HSR is a redundancy protocol for Ethernet networks that provides duplicated frames for separate physical paths with zero recovery time. This feature makes it very useful for real-time and mission-critical applications, such as military applications and substation automation systems. However, the major drawback of HSR is that it generates too much unnecessary redundant traffic in HSR networks. This drawback degrades network performance and may cause congestion and delay. Several HSR traffic reduction techniques have been proposed to reduce the redundant traffic in HSR networks, resulting in the improvement of network performance. In this paper, we provide an overview of these HSR traffic reduction techniques in the literature. The operational principles, advantages, and disadvantages of these techniques are investigated and summarized. We also provide a traffic performance comparison of these HSR traffic reduction techniques.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.12
• /
• pp.1665-1670
• /
• 2018

A convolutional neural network (CNN), which is one of the deep learning models, has been very successful in a variety of computer vision tasks. Filters of a CNN are automatically generated, however, they can be further optimized since there exist the possibility of existing redundant and less important features. Therefore, the aim of this paper is a filter reduction to accelerate and compress CNN models. Evolutionary algorithms is adopted to remove the unnecessary filters in order to minimize the parameters of CNN networks while maintaining a good performance of classification. We demonstrate the proposed filter reduction methods performing experiments on CIFAR10 data based on the classification performance. The comparison for three approaches is analysed and the outlook for the potential next steps is suggested.

• International Journal of Computer Science & Network Security
• /
• v.21 no.3
• /
• pp.31-36
• /
• 2021

RNA sequencing (RNA-Seq) is a technology that facilitates transcriptome analysis using next-generation sequencing (NSG) tools. Information on the quantity and sequences of RNA is vital to relate our genomes to functional protein expression. RNA-Seq data are characterized as being high-dimensional in that the number of variables (i.e., transcripts) far exceeds the number of observations (e.g., experiments). Given the wide range of dimensionality reduction techniques, it is not clear which is best for RNA-Seq data analysis. In this paper, we study the effect of three dimensionality reduction techniques to improve the classification of the RNA-Seq dataset. In particular, we use PCA, SVD, and SOM to obtain a reduced feature space. We built nine classification models for a cancer dataset and compared their performance. Our experimental results indicate that better classification performance is obtained with PCA and SOM. Overall, the combinations PCA+KNN, SOM+RF, and SOM+KNN produce preferred results.

• Journal of Contemporary Eastern Asia
• /
• v.18 no.2
• /
• pp.90-117
• /
• 2019

How is science advice integrated in environmental policymaking? This is an increasingly pertinent question that is being raised since the nuclear catastrophe of Fukushima, Japan, in 2011. Global re-evaluation of energy policies and climate mitigation measures include discussions on how to better integrate science advice in policymaking, and at the same time keeping science independent from political influence. This paper addressed the policy discourse of setting up a national CO2 reduction target in Japanese policymaking between 2009 and 2012. The target proposed by the former DPJ government was turned down, and Japan lacked a clear strategy for long-term climate mitigation. The analysis provides explanations from a quantitative actor-network perspective. Centrality measures from social network analysis for policy actors in an environmental policy network of Japan were calculated to identify those actors that control the discourse. Data used for analysis comes from the Global Environmental Policy Actor Network 2 (GEPON 2) survey conducted in Japan (2012-13). Science advice in Japan was kept independent from political influence and was mostly excluded from policymaking. One of the two largest discourse coalitions in the environmental policy network promoted a higher CO2 reduction target for international negotiations but favored lowering the target after a new international agreement would have been set. This may explain why Japan struggled to commit to long-term mitigation strategies. Applying social network analysis to quantitatively calculate discourse coalitions was a feasible methodology for investigating "discursive power." But limited in discussing the "practice" (e.g. meetings, telephone, or email conversations) among the actors in discourse coalitions.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.9 no.4
• /
• pp.301-308
• /
• 2009

In the modern life, machines are used for various areas in industries as the advance of science and industrial development has proceeded. In many machines, the rotating machines play an important role in many processes. Therefore, the development of fault diagnosis and monitoring system for rotating machines is required. An ubiquitous sensor network (USN) is a combination of the key computer science and engineering area technology including the wireless network, embedded system hardware and software, communication, real-time system, etc. It collects environmental information to realize a variety of functions. In this work, a data reduction algorithm for USN based remote fault diagnostic system which can be easily applied to previously built factories is proposed. To verify the feasibility of the proposed scheme, some simulations and experiments are executed.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.19 no.2
• /
• pp.57-68
• /
• 2023

Because salt and pepper noise is a type of impulse, even a small amount of noise could cause a large image degradation. In this paper, we proposed a salt-and-pepper noise removal method using the convolutional neural network. It consists of four phases. In the first step, the proposed method reconstructs noisy image using a traditional salt-and-pepper noise reduction method, and in the second step, the result image of previous step is filtered with Gaussian low pass filter. After that, we reconstruct the filtered image using convolution neural network. In the last step, the pixels with salt-and-pepper noise are replaced with the result of previous phase. Simulation results show that the proposed method yields not only objective image qualities(PSNR, SSIM) but also subjective image qualities for all SAP noise ratios.