• Electronics and Telecommunications Trends
• /
• v.34 no.4
• /
• pp.32-42
• /
• 2019

The performance of machine learning algorithms significantly depends on how a configuration of hyperparameters is identified and how a neural network architecture is designed. However, this requires expert knowledge of relevant task domains and a prohibitive computation time. To optimize these two processes using minimal effort, many studies have investigated automated machine learning in recent years. This paper reviews the conventional random, grid, and Bayesian methods for hyperparameter optimization (HPO) and addresses its recent approaches, which speeds up the identification of the best set of hyperparameters. We further investigate existing neural architecture search (NAS) techniques based on evolutionary algorithms, reinforcement learning, and gradient derivatives and analyze their theoretical characteristics and performance results. Moreover, future research directions and challenges in HPO and NAS are described.

• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2023.11a
• /
• pp.261-262
• /
• 2023

Drone is a suitable equipment for capturing images of cracks at construction sites based on its efficient mobility and high-resolution image acquisition capabilities. In this study, drone was used to acquire indoor construction sites framework images and deep learning technology was applied to detect cracks and measure width, and size. Finally, the usability of the process was verified based on the indoor crack monitoring system.

• Journal of the Semiconductor & Display Technology
• /
• v.23 no.2
• /
• pp.28-32
• /
• 2024

The personal mobility Sharing service is currently active. Especially, electric scooters are widely utilized because they can move comfortably at a high speed over a short distance with a simple driving method. Its driving method is easy, but there is no protection device to protect the bare body. So, there is a greater accident than other means of transportation, and if two people are on board, there is higher accident probability. However, since there is no specific ways to prevent multi-person boarding yet, we propose a multi-person boarding detection model using tilt and pressure sensor. The proposed method measures the tilt degree and direction by using a tilt sensor installed in the center of the board plate and detects multi-people riding.

• Journal of Auto-vehicle Safety Association
• /
• v.16 no.2
• /
• pp.60-66
• /
• 2024

Computer simulation significantly reduces the high costs associated with actual crash tests and is expanding due to its ability to analyze various test results quantitatively that are difficult to measure in real tests. Research on evaluation technologies is limited according to the finite element analysis, which aims to replace structural verification testing. In this study, considerations for virtual crash tests were derived, and the validity of the zero-energy mode (hourglass mode) was analyzed as part of the considerations for validating the results of vehicle crash simulations. The study reflects on the considerations for virtual crash tests and the variation in hourglass coefficient values affects the occurrence of the hourglass mode. As the hourglass coefficient changes, the maximum hourglass energy reaches over 5% of the maximum internal energy, necessitating a conservative review. A comprehensive study of the maximum hourglass energy is required, considering additional analysis results for various models and collision conditions.

• Fashion & Textile Research Journal
• /
• v.10 no.1
• /
• pp.1-10
• /
• 2008

• Journal of Technology Innovation
• /
• v.26 no.2
• /
• pp.95-128
• /
• 2018

Since mid and late 2000s, the smartphones has been widely diffused and Korea ranked first in global smartphone market in 2011 thanks to its rapid adoption of Android operating system, technology capability accumulated in featurephone development, vertical integration on smartphone production and premium positioning. However, Korea fell behind because of the rise of another latecomer, China, in four years (2015) after it recorded the top position globally. How did the leadership change occurred in the smartphone industry so rapidly? In order to answer the question, we investigated three favorable windows of opportunity for the rise of China, which are technological, demand, and institutional, and the strategic responses of Chinese firms as well as the rigidity and complacency with the past success of Korean firms. Our findings contribute to the extension of 'catch-up cycle' theory as well as provide in-depth insights for strategies and policies settings to overcome the recent rise of China in information and communication technology sector for Korea.

• Journal of Internet Computing and Services
• /
• v.14 no.3
• /
• pp.47-55
• /
• 2013

In recent years, smart devices have changed the paradigm of education. However, the educational environment and teaching methods could not catch up with this fast improvement and an utmost need for development of educational methods has been realized. In this paper, the general usage of smart devices by elementary school teachers is analyzed and the methods through which smart devices are utilizing smart learning is discussed in the result. A survey of 221 elementary school teachers in Gangwon Province showed that 87.7% of them are currently using smart devices. In addition, teachers were using smart devices with not much difference from ordinary people. The three main motives of teachers who were already using smart devices and those who were planning to take advantage of smart devices were innovativeness, usability and easiness. The reason of need to apply smart devices in education is because of its functionality in various learning types, courses and teaching-learning process methods. Although smart devices have a high efficacy in education but they are not widely utilized yet. In order to solve these problems and be able to take more advantage of smart devices in education, teachers should learn how to use smart devices and a strong sense of willingness is required to make changes in the educational methods. The results of this research on elementary school teachers can be further developed for a greater smart device based smart learning.

• Journal of Technology Innovation
• /
• v.28 no.4
• /
• pp.97-126
• /
• 2020

This study evaluated its potentials by quantitatively analyzing the national economic impact of the smart home-related industry, which is attracting attention as a core industry of the 4th industrial revolution. For the analysis, the smart home-related industries were classified into manufacturing and service industries through a literature review of the previous studies. Using the 2018 input-output table, this paper analyzed linkage effects between industries as well as spillover effects in the production, value-added, employment and job. As a result, the smart home manufacturing and service sectors showed a higher spillover effect in value-added than other industries in each industrial field. In the smart home industry, the spillover effects of manufacturing sector to service sector are larger than those of service sector to manufacturing sector. Moreover, it was confirmed that smart home industry was highly related to not only the technology-intensive industry, but also the service sector for smart cities, smart cars, Fin-tech, and etc. On the other hand, the smart home manufacturing sector is a final demanding industry with relatively higher backward linkage effect than forward linkage effect. In the smart home service sector, the forward linkage effect was relatively high compared with the backward linkage effect, indicating that it was an industry with a high supply function to other industries.

• Journal of Digital Forensics
• /
• v.12 no.3
• /
• pp.109-120
• /
• 2018

The smartphone is the communication device that is the most personal to the user, and it keeps a lot of information related to the user and makes information communication with other devices. With these characteristics, forensics on smartphones are one of the most basic methods of investigation in criminal investigations, and have actually contributed to the settlement of the case by providing many clues. However, recently, it is designed to encrypt data stored as a social issue related to the protection of user's personal information, or to delete deleted data or to delete log data together. So, any solutions? In this paper, I try to find the answer from cloud data stored by smartphone user account. Cloud forensics should approach complementary relationships rather than smartphone forensics. There are a lot of data stored in the cloud that can be meaningfully used in the investigation. Online activity information of users, such as Internet usage, YouTube view, and contents purchase information, cloud service such as e-mail, cloud drive, and location information are the most representative data. These data can be unvaluable, but here are some important clues in various types of criminal investigations. In this paper, I propose a method to extract data from the google cloud so that the data can be used for investigation, and to utilize the extracted data for investigation. And it explains the role of the extracted artifacts in the actual investigation business through virtual cases and proves its value.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.18 no.1
• /
• pp.74-82
• /
• 2019

The revolution of industry 4.0 is enabling us to build an intelligent connection society called smart cities. The use of renewable energy in particular solar energy is extremely important for modern society due to the growing power demand in smart cities, but its difficult to monitor and manage in each buildings since need to be deploy low energy sensors and information need to be transfer via wireless sensor network (WSN). The Internet of Things (IoT) / low-power wide-area (LPWA) is an emerging WSN technology, to collect and monitor data about environmental and physical electrical / electronics devices conditions in real time. However, providing power to IoT sensor end devices and other public electrical loads such as street lights, etc is an important challenging role because the sensor are usually battery powered and have a limited life time. In this paper, we proposes an efficient solar energy-based power management scheme for smart city based on IoT technology using LoRa wide-area network (LoRaWAN). This approach facilitates to maintain and prevent errors of solar panel based energy systems. The proposed solution maximizing output the power generated from solar panels system to distribute the power to the load and the grid. In this paper, we proved the efficiency of the proposed system with Simulink based system modeling and real-time emulation.