• The Korean Journal of Applied Statistics
• /
• v.32 no.5
• /
• pp.693-702
• /
• 2019

Deep learning has gained popularity for the classification and prediction task. Neural network layers become deeper as more data becomes available. Saturation is the phenomenon that the gradient of an activation function gets closer to 0 and can happen when the value of weight is too big. Increased importance has been placed on the issue of saturation which limits the ability of weight to learn. To resolve this problem, Glorot and Bengio (Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, 249-256, 2010) claimed that efficient neural network training is possible when data flows variously between layers. They argued that variance over the output of each layer and variance over input of each layer are equal. They proposed a method of initialization that the variance of the output of each layer and the variance of the input should be the same. In this paper, we propose a new method of establishing initialization by adopting truncated normal distribution and truncated cauchy distribution. We decide where to truncate the distribution while adapting the initialization method by Glorot and Bengio (2010). Variances are made over output and input equal that are then accomplished by setting variances equal to the variance of truncated distribution. It manipulates the distribution so that the initial values of weights would not grow so large and with values that simultaneously get close to zero. To compare the performance of our proposed method with existing methods, we conducted experiments on MNIST and CIFAR-10 data using DNN and CNN. Our proposed method outperformed existing methods in terms of accuracy.

• International Journal of Highway Engineering
• /
• v.19 no.4
• /
• pp.1-7
• /
• 2017

PURPOSES : This study aims to improve complex modeling of multivariable, nonlinear, and overdispersion data with an artificial neural network that has been a problem in the civil and transport sectors. METHODS: Deep learning, which is a technique employing artificial neural networks, was applied for developing a large bus fuel consumption model as a case study. Estimation characteristics and accuracy were compared with the results of conventional multiple regression modeling. RESULTS : The deep learning model remarkably improved estimation accuracy of regression modeling, from R-sq. 18.76% to 72.22%. In addition, it was very flexible in reflecting large variance and complex relationships between dependent and independent variables. CONCLUSIONS : Deep learning could be a new alternative that solves general problems inherent in conventional statistical methods and it is highly promising in planning and optimizing issues in the civil and transport sectors. Extended applications to other fields, such as pavement management, structure safety, operation of intelligent transport systems, and traffic noise estimation are highly recommended.

• Journal of Korean Clinical Health Science
• /
• v.6 no.2
• /
• pp.1136-1146
• /
• 2018

purpose: The study was performed by targeting nursing students in order to examine the level of their self-directed learning ability, self efficacy and learner satisfaction and to identify their effects on learner satisfaction. methods: A structured self-administered questionnaire was used, and a total of 150 questionnaires were distribute. Data were collected from 150 nursing students at D university located in D city, and a total of 141 students were included for the final analysis. Data were collected between November 15, 2017 to December 16, 2017, and questionnaire comprised items to measure general characteristics, self-directed learning ability, self efficacy and learner satisfaction. results: In addition, learner satisfaction was positively correlated with self-direct learning ability, self-efficacy. Finally, self-direct learning ability, self-efficacy and interpersonal relationship were found to explain 42.9% of the variance of learner satisfaction. conclusion: The findings of this study could be utilized as base data when developing a program to enhance nursing students's learner satisfaction and it is suggested that an interventional research of analyzing validity and effectiveness of developed education program would be also required.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.14 no.6
• /
• pp.1153-1160
• /
• 2019

In this paper, we experimented to find out how the number of convolution layers, the size, and the number of kernels affect the CNN. In addition, the general CNN was also tested for analysis and compared with the CNN used in the experiment. The neural networks used for the analysis are based on CNN, and each experimental model is experimented with the number of layers, the size, and the number of kernels at a constant value. All experiments were conducted using two layers of fully connected layers as a fixed. All other variables were tested with the same value. As the result of the analysis, when the number of layers is small, the data variance value is small regardless of the size and number of kernels, showing a solid accuracy. As the number of layers increases, the accuracy increases, but from above a certain number, the accuracy decreases, and the variance value also increases, resulting in a large accuracy deviation. The number of kernels had a greater effect on learning speed than other variables.

• Smart Structures and Systems
• /
• v.22 no.2
• /
• pp.231-237
• /
• 2018

A significant data problem is encountered with condition monitoring because the sensors need to measure vibration data at a continuous and sometimes high sampling rate. In this study, compressive sensing approaches for condition monitoring are proposed to demonstrate their efficiency in handling a large amount of data and to improve the damage detection capability of the current condition monitoring process. Compressive sensing is a novel sensing/sampling paradigm that takes much fewer data than traditional data sampling methods. This sensing paradigm is applied to condition monitoring with an improved machine learning algorithm in this study. For the experiments, a built-in rotating system was used, and all data were compressively sampled to obtain compressed data. The optimal signal features were then selected without the signal reconstruction process. For damage classification, we used the Variance Considered Machine, utilizing only the compressed data. The experimental results show that the proposed compressive sensing method could effectively improve the data processing speed and the accuracy of condition monitoring of rotating systems.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.17 no.4
• /
• pp.703-708
• /
• 2022

K-means clustering is a representative clustering technique. However, there is a limitation in not being able to integrate the performance evaluation scale and the method of determining the minimum number of clusters. In this paper, a method for numerically determining the minimum number of clusters is introduced. The explained variance is presented as an integrated measure. We propose that the k-means clustering method should be performed in the subspace of the PCA in order to simultaneously satisfy the minimum number of clusters and the threshold of the explained variance. It aims to present an explanation in principle why principal component analysis and k-means clustering are sequentially performed in pattern recognition and machine learning.

• Journal of Korean Society of Archives and Records Management
• /
• v.22 no.1
• /
• pp.61-82
• /
• 2022

This study aims to investigate the factors influencing the learning satisfaction of records management cyber education that is opened and operated by the National Archives of Korea, as well as to improve the quality of cyber education program. Cyber education consisted of an introductory course, an intensive course, and a liberal arts course. As the major factors for learning satisfaction, "validity of content structure, interaction between professors and learners, learning motivation, active learning attitude, ease of use environment, and level of organizational support" were set. An online survey was conducted on the learning satisfaction according to the curriculum of each course. The survey was conducted to 107 institutions with specialized records management personnel, and additional in-depth interviews were also conducted. The survey analysis consisted of factor analysis, independent sample T-test, analysis of variance (ANOVA), correlation analysis, and multiple regression analysis. As a result of the study, the factors influencing learning satisfaction were found in the order of interaction between professors and learners, learning motivation, and validity of content structure.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.11a
• /
• pp.305-309
• /
• 2000

In this paper, we presents a new methodology for face recognition after analysing conventional ICA(Independent Component Analysis) based approach. In the literature we found that ICA based methods have followed the same procedure without any exception, first PCA(Principal Component Analysis) has been used for feature extraction, next ICA learning method has been applied for feature enhancement in the reduced dimension. However, it is contradiction that features are extracted using higher order moments depend on variance, the second order statistics. It is not considered that a necessary component can be located in the discarded feature space. In the new methodology, features are extracted using the magnitude of kurtosis(4-th order central moment or cumulant). This corresponds to the PCA based feature extraction using eigenvalue(2nd order central moment or variance). The synergy effect of PCA and ICA can be achieved if PCA is used for noise reduction filter. ICA methodology is analysed using SVD(Singular Value Decomposition). PCA does whitening and noise reduction. ICA performs the feature extraction. Simulation results show the effectiveness of the methodology compared to the conventional ICA approach.

• Korean Journal of Child Studies
• /
• v.30 no.4
• /
• pp.87-97
• /
• 2009

This study used Hierarchical Linear Modeling analysis to investigate the effects of individual self-regulated cognitive strategies and public education on middle school students' academic achievement. Participants were 6389 (boys 3287, girls 3102) middle school students from the 2005 data of the Korea Education Longitudinal Study. Results were as follows : (1) there were significant differences among different schools in middle school students' academic achievement, i.e. 20% of variance in English achievement and 15% of variance in mathematics achievement were explained by school differences. (2) Students' elaboration and meta-cognitive strategy influenced academic achievement positively. (3) Predictor variables by ability grouping, supplementary class, and/or self-learning class had no significant effects on students' academic achievement.

• Journal of the Korean Data and Information Science Society
• /
• v.15 no.3
• /
• pp.523-529
• /
• 2004

In Korea, mathematics education has been changed according to the 7th national mathematics curriculum renovated by the Ministry of Education and Human Resources Development announcement in 1997. The education of probability and Statistics has been carried out as a part of this curriculum. We analyze and compare 3 kinds of mathematics textbooks for 10-12 grade students. Descriptions of random variable, sample variance and sample standard deviation, distribution of sample mean, and etc. which are on some textbooks, are misleaded in school education. We suggest the unbiased estimator of sample variance in textbooks and distributions of sample means with normal population assumption.