• The Korean Journal of Applied Statistics
• /
• v.32 no.2
• /
• pp.199-211
• /
• 2019

This paper is concerned with issues in the finite mixture of regression modeling as well as the simultaneous selection of the number of mixing components and relevant predictors. We propose a penalized likelihood method for both mixture components and regression coefficients that enable the simultaneous identification of significant variables and the determination of important mixture components in mixture of regression models. To avoid over-fitting and bias problems, we applied smoothly clipped absolute deviation (SCAD) penalties on the logarithm of component probabilities suggested by Huang et al. (Statistical Sinica, 27, 147-169, 2013) as well as several well-known penalty functions for coefficients in regression models. Simulation studies reveal that our method is satisfactory with well-known penalties such as SCAD, MCP, and adaptive lasso.

• Korean Journal of Optics and Photonics
• /
• v.30 no.2
• /
• pp.48-58
• /
• 2019

We investigate the signal-to-noise ratios (SNRs) of time-domain (i.e. pulsed illumination) and frequency-domain (i.e. chirped illumination) photoacoustic signals measured by a spherically focused ultrasound transducer for spherical absorbers. The simulation results show that the time-domain photoacoustic SNR is higher than that of frequency-domain photoacoustic signals, as reported in the previous literature. We understand the reason for this SNR gap between the two measurement modes by analyzing photoacoustic-signal spectra, considering the incident beam energy controlled by the maximum permissible exposure. As the result of this approach, we find that filtering off the DC term in the chirped signal's spectrum improves frequency-domain photoacoustic SNRs by up to approximately 5 dB. In particular, it is observed that photoacoustic SNRs are highly sensitive to an upper-frequency value of frequency filtering functions, and the optimal upper-frequency values maximizing the SNR are different in time- and frequency-domain photoacoustic measurements.

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.1
• /
• pp.78.2-78.2
• /
• 2019

We develop an Alcock-Paczynski (AP) test method that uses the evolution of redshift-space two-point correlation function (2pCF) of galaxies. The method improves the AP test proposed by Li et al. (2015) in that it uses the full two-dimensional shape of the correlation function. Similarly to the original method, the new one uses the 2pCF in redshift space with its amplitude normalized. Cosmological constraints can be obtained by examining the redshift dependence of the normalized 2pCF. This is because the 2pCF should not change apart from the expected small non-linear evolution if galaxy clustering is not distorted by incorrect choice of cosmology used to convert redshift to comoving distance. Our new method decomposes the redshift difference of the 2-dimensional correlation function into the Legendre polynomials whose amplitudes are modelled by radial fitting functions. The shape of the normalized 2pCF suffers from small intrinsic time evolution due to non-linear gravitational evolution and change of type of galaxies between different redshifts. It can be accurately measured by using state of the art cosmological simulations. We use a set of our Multiverse simulations to find that the systematic effects on the shape of the normalized 2pCF are quite insensitive to change of cosmology over \Omega_m=0.21 - 0.31 and w=-0.5 - -1.5. Thanks to this finding, we can now apply our method for the AP test using the non-linear systematics measured from a single simulation of the fiducial cosmological model.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.4
• /
• pp.394-400
• /
• 2019

Noise removal is one of the required step in processing digital video and there are many researches to develop algorithm that fits with its purpose and environment. However, present impulse noise removal methods are lacking in its function in terms of removing noise in edge and high frequency factors. Therefore, this research has Extended range of masks depending on density to determine noise so that high frequency factors can be preserved. The range of resolution is set based on median and standard deviation of inside resolution after removing impulse noise. afterwards, those resolution within the range are calculated by adding weight to have the final output value. The suggested algorithm has an enhanced function in removing noise in various areas with many edge and high frequency factors than present methods and their functions are compared through simulation.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.9 no.1
• /
• pp.179-187
• /
• 2019

With the recent development of small computing devices, IoT sensor network can be widely deployed and is now readily available with sensing, calculation and communi-cation functions at low cost. Sensor data management is a major component of the Internet of Things environment. The huge volume of data produced and transmitted from sensing devices can provide a lot of useful information but is often considered the next big data for businesses. New column-wise compression technology is mounted to the large data server because of its superior space efficiency. Since sensor nodes have narrow bandwidth and fault-prone wireless channels, sensor-based storage systems are subject to incomplete data services. In this study, we will bring forth a short overview through providing an analysis on IoT sensor networks, and will propose a new storage management scheme for IoT data. Our management scheme is based on RAID storage model using column-wise segmentation and compression to improve space efficiency without sacrificing I/O performance. We conclude that proposed storage control scheme outperforms the previous RAID control by computer performance simulation.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.1
• /
• pp.131-139
• /
• 2021

It can be said that smart factory is the most prominent area in the fourth industrial revolution. Developing processes or algorithms required for smart factory requires data values from smart factory, but there are many real challenges in obtaining such data. Therefore, this study developed a data generator that can more realistically simulate data from different processes in smart factory to help research on smart factory. In addition, functions such as setting presets and intuitive UI configurations were developed for the convenience of data creators. This data generator will help you simulate smart factory environments by providing more realistic data easily and simply when you create the different systems needed for smart factory environments.

• The Korean Journal of Applied Statistics
• /
• v.34 no.3
• /
• pp.411-425
• /
• 2021

Accelerated failure time (AFT) model represents a linear relationship between the log-survival time and covariates. We are interested in the inference of covariate's effect affecting the variation of survival times in the AFT model. Thus, we need to model the variance as well as the mean of survival times. We call the resulting model mean and variance AFT (MV-AFT) model. In this paper, we propose a variable selection procedure of regression parameters of mean and variance in MV-AFT model using penalized likelihood function. For the variable selection, we study four penalty functions, i.e. least absolute shrinkage and selection operator (LASSO), adaptive lasso (ALASSO), smoothly clipped absolute deviation (SCAD) and hierarchical likelihood (HL). With this procedure we can select important covariates and estimate the regression parameters at the same time. The performance of the proposed method is evaluated using simulation studies. The proposed method is illustrated with a clinical example dataset.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.1
• /
• pp.25-36
• /
• 2021

This paper presents a topology based on packet tracer and a five-step scenario model to make it easier for students to understand the network on which VLANs are applied. Virtual LAN (VLAN), developed as an alternative solution to the Routers that distribute broadcast traffic, is a virtual local area network that logically configured regardless of the physical network. The VLAN prevents the network performance degradation resulting from the broadcast traffic by dividing the broadcast domain so that the bandwidth can be used more efficiently. In addition, it enhances the security because on communication between the devices belonging to different VLANs is impossible. The five-step scenarios in this paper presented an efficient implementation case for students to understand and validate the various functions of VLANs through ping/telnet/tracert test and simulation, after setting up each step of programming switches and routers in the virtual network.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.24 no.10
• /
• pp.1319-1324
• /
• 2020

A smart tag complements various problems caused by the expansion of the use of existing magnet cards. This emerged as a result of the need for alternative means to perform safer and more diverse functions. RFID tags are becoming increasingly ubiquitous sensor networks and Internet of Things technologies as they give everything an ID, enabling automatic recognition of objects and forming a communication network between them. In this paper, a contactless RFID smart tag reader composed of RFID tags and readers using AVR processor was developed. The system has a low frequency band of 13.56MHz and is designed by PWM method of communication velocity 26.48kbps. The developed RFID smart tag board showed a faster response rate than the typical RFID system as a result of the simulation, and the actual experiment also showed an improvement in response rate.

• Steel and Composite Structures
• /
• v.39 no.1
• /
• pp.51-64
• /
• 2021

This paper presents a high-order shear and normal deformation theory for the bending of FGM plates. The number of unknowns and governing equations of the present theory is reduced, and hence makes it simple to use. Unlike any other theory, the number of unknown functions involved in displacement field is only four, as against five or more in the case of other shear and normal deformation theories. Based on the novel shear and normal deformation theory, the position of neutral surface is determined and the governing equilibrium equations based on neutral surface are derived. There is no stretching-bending coupling effect in the neutral surface-based formulation, and consequently, the governing equations of functionally graded plates based on neutral surface have the simple forms as those of isotropic plates. Navier-type analytical solution is obtained for functionally graded plate subjected to transverse load for simply supported boundary conditions. The accuracy of the present theory is verified by comparing the obtained results with other quasi-3D higher-order theories reported in the literature. Other numerical examples are also presented to show the influences of the volume fraction distribution, geometrical parameters and power law index on the bending responses of the FGM plates are studied.