• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.4
• /
• pp.639-644
• /
• 2002

Nonlinear filters which are utilized rank-order information and temporal-order information, have many proposed, in order to restore nonstationary signals which are corrupted by additive noise. In this paper, we propose a data-dependent LWOS filter whose coefficients change based on local statistics. LWOS(Linear Combination of Weighted Order Statistics) filters［1］which also utilized two informations, and have properties of efficient impulsive and nonimpulsive noise attenuation and sufficiently details and edges preservation. DD-LWOS filters can remove non-impulsive oises while preserving signal details. DD-LWOS2 filter gets more better performance than DD-LWOS filter when input image corrupted by additive noise which includes Impulsive noise components.

• The Korean Journal of Applied Statistics
• /
• v.29 no.4
• /
• pp.627-641
• /
• 2016

In this paper, we consider a seemingly unrelated regression (SUR) model and propose a nonparametric Bayesian approach to SUR with a Dirichlet process mixture of normals for modeling an unknown error distribution. Posterior distributions are derived based on the proposed model, and the posterior inference is performed via Markov chain Monte Carlo methods based on the collapsed Gibbs sampler of a Dirichlet process mixture model. We present a simulation study to assess the performance of the model. We also apply the model to precipitation data over South Korea.

• 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.

• Journal of the Korean Society for Library and Information Science
• /
• v.41 no.1
• /
• pp.431-451
• /
• 2007

As a new practical model of Korean Library statistics, It suggests 6 parts, fifty-four items for 4 types of libraries. In comparison with last version, 'library use and users', 'electronic services' parts are increased which are occupied as more than hair of the whole statistical items. Budget parts are increased also. The new model which is developed based on the international standards like ISO2789 and NISO/ANSI Z39.7 is advised by professionals in the fields.

• Korean Journal of Adult Nursing
• /
• v.24 no.1
• /
• pp.20-30
• /
• 2012

Purpose: This study was to explore the gap between awareness and performance toward evidence-based pain management in older adults for the purpose of establishing baseline data for evidence-based pain management protocol development and dissemination. Methods: The subjects were 290 staff nurses from three general hospitals. Self administered questionnaires were used to collect the data and the results of the study were analyzed with descriptive statistics, t-test, ANOVA and Pearson's correlation. Results: There were statistically significant differences between awareness and performance in pain assessment (t=17.31, $p$ <.001), patient and family education (t=17.33, $p$ <.001), pharmacologic management (t=12.99, $p$ <.001), non pharmacological management (t=16.28, p<.001), and evaluation of effectiveness (t=11.70, $p$ <.001). There were also statistically significant differences in awareness and performance according to the workplace, knowledge, and usual performance. Conclusion: The study showed that the hospital nurses' performance about evidence-based pain management in older adults was lower than their awareness level thus indicating significant gaps between evidence and actual practice. To ensure effective pain care, the factors contributing to these gaps need to be analyzed to identify the barriers. In addition, the evidence-based pain management guideline suitable for various clinical settings needs to be developed and disseminated.

• Journal of the Korean Data and Information Science Society
• /
• v.26 no.2
• /
• pp.517-524
• /
• 2015

Unlabeled examples are easier and less expensive to be obtained than labeled examples. In this paper semisupervised approach is used to utilize such examples in an effort to enhance the predictive performance of nonlinear quantile regression problems. We propose a semisupervised quantile regression method named semisupervised support vector quantile regression, which is based on support vector machine. A generalized approximate cross validation method is used to choose the hyper-parameters that affect the performance of estimator. The experimental results confirm the successful performance of the proposed S2SVQR.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.53 no.9
• /
• pp.663-669
• /
• 2004

The mobile tole-cardiology system is the new research area that support an ubiquitous health care based on mobile telecommunication networks. Although there are many researches presenting the modeling concepts of a GSM-based mobile telemedical system, practical application needs to be considered both compression performance and error corruption in the mobile environment. This paper evaluates three wavelet ECG compression algorithms over CDMA networks. The three selected methods are Rajoub using EPE thresholding, Embedded Zerotree Wavelet(EZW) and Wavelet transform Higher Order Statistics Coding(WHOSC) with linear prediction. All methodologies protected more significant information using Forward Error Correction coding and measured not only compression performance in noise-free but also error robustness and delay profile in CDMA environment. In addition, from the field test we analyzed the PRD for movement speed and the features of CDMA 1X. The test results show that Rajoub has low robustness over high error attack and EZW contributes to more efficient exploitation in variable bandwidth and high error. WHOSC has high robustness in overall BER but loses performance about particular abnormal ECG.

• Journal of Korean Clinical Nursing Research
• /
• v.29 no.2
• /
• pp.149-162
• /
• 2023

Purpose: This study was to investigate the belief in evidence-based practice, awareness of importance and performance of intravenous infusion and pressure ulcer evidence-based practice guidelines among nurses in a tertiary general hospital. Methods: The subjects of this study were 217 nurses working in a tertiary general hospital. Data collection was performed between February 11 and February 25, 2022. Data analysis was conducted descriptive statistics, t-test, hierarchical regression analysis, and Importance-Performance Analysis. Results: The mean score of belief for evidence-based practice among novice nurses was 3.34 out of 5, while preceptor nurses scored a mean of 3.41 out of 5. There was no significant difference in belief scores between novice nurses and preceptor nurses (t=-1.21, p=.227). The factors influencing the performance of evidence-based practice guidelines for intravenous infusion were belief in evidence-based practice (β=.14, p=.009) and importance of intravenous infusion (β=.51, p<.001), and the factors influencing the performance of evidence-based practice guidelines for pressure ulcer were belief in evidence-based practice (β=.15, p=.002) and importance of pressure ulcer (β=.65, p<.001). Importance-Performance Analysis of the evidence-based practice guidelines of two groups were used to identify common and different items. Conclusion: To improve the performance of evidence-based practice guidelines, it is necessary to enhance the evidence-based practice belief and importance of evidence-based practice guidelines. In particular, evidence-based practice should be provided to improve nursing quality through education on items of low-importance and low-performance and items of high-importance but low-performance guidelines identified through Importance-Performance Analysis.

• The Korean Journal of Applied Statistics
• /
• v.37 no.2
• /
• pp.139-155
• /
• 2024

This study performed a multi-classification task to classify mesial temporal lobe epilepsy with left hippocampal sclerosis patients (left mTLE), mesial temporal lobe epilepsy with right hippocampal sclerosis (right mTLE), and healthy controls (HC) using magnetoencephalography (MEG) data. We applied various artificial neural networks and compared the results. As a result of modeling with convolutional neural networks (CNN), recurrent neural networks (RNN), and graph neural networks (GNN), the average k-fold accuracy was excellent in the order of CNN-based model, GNN-based model, and RNN-based model. The wall time was excellent in the order of RNN-based model, GNN-based model, and CNN-based model. The graph neural network, which shows good figures in accuracy, performance, and time, and has excellent scalability of network data, is the most suitable model for brain research in the future.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.45 no.6
• /
• pp.148-155
• /
• 2008

In this paper, we constructed the noise robust content-based music retrieval system in mobile environment. The performance of the proposed system was verified with ZCPA feature which is blown to have noise robust characteristic in speech recognition application. In addition, new indexing and fast retrieval method are proposed to improve retrieval speed about 99% compare to exhaustive retrieval for large music DB. From the computer simulation results in noise environment of 15dB - 0dB SNR, we confirm the superior performance of the proposed system about 5% - 30% compared to MFCC and FBE(filter bank energy) feature.