• Journal of Internet of Things and Convergence
• /
• v.8 no.5
• /
• pp.77-83
• /
• 2022

Nand flash memory requires an Erase-Before-Write operation structurally. In order to solve this problem, it can be solved by classifying a page (hot data page) where data update operation occurs frequently and storing it in a separate block. The MHF (Multi Hash Function Framework) technique records the frequency of data update requests in the system memory, and when the recorded value exceeds a certain standard, the data update request is judged as hot data. However, the method of simply counting only the frequency of the data update request has a limit in judging it as accurate hot data. In addition, in the case of a technique that determines the persistence of a data update request, the fact of the update request is recorded sequentially based on a time interval and then judged as hot data. In the case of such a persistence-based method, its implementation and operation are complicated, and there is a problem of inaccurate judgment if frequency is not considered in the update request. This paper proposes a lightweight hot data determination technique that considers both frequency and persistence in data update requests.

• Journal of Internet of Things and Convergence
• /
• v.8 no.5
• /
• pp.33-39
• /
• 2022

A storage device used in the mobile computing field should have low power, light weight, durability, etc., and should be able to effectively store and manage large-capacity data generated by users. NAND flash memory is mainly used as a storage device in the field of mobile computing. Due to the structural characteristics of NAND flash memory, it is impossible to overwrite in place when a data update request is made, so it can be solved by accurately separating requests that frequently request data update and requests that do not, and storing and managing them in each block. The classification method for such a data update request is called a hot data identification method, and various studies have been conducted at present. This paper continuously records the occurrence of data update requests using a counting filter for more accurate hot data validation, and also verifies hot data by considering how often the requested update requests occur during a specific time.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.25 no.5
• /
• pp.135-140
• /
• 2021

In this study, the concrete aggregate shape features were extracted from the cross-section of a normal concrete strength cylinder, and the compressive strength of the cylinder was predicted using artificial neural networks and image processing technology. The distance-angle features of aggregates, along with general aggregate shape features such as area, perimeter, major/minor axis lengths, etc., were numerically expressed and utilized for the compressive strength prediction. The results showed that compressive strength can be predicted using only the aggregate shape features of the cross-section without using major variables. The artificial neural network algorithm was able to predict concrete compressive strength within a range of 4.43% relative error between the predicted strength and test results. This experimental study indicates that various material properties such as rheology, and tensile strength of concrete can be predicted by utilizing aggregate shape features.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.15 no.6
• /
• pp.492-505
• /
• 2022

This study is a study on the results of improving the logic to effectively transmit and decode compressed data in real time by improving the encoding and decoding performance of BL-beta codes that can be used for lossless real-time transmission of IoT sensing data. The encoding process of BL-beta code includes log function, exponential function, division and square root operation, etc., which have relatively high computational burden. To improve them, using bit operation, binary number pattern analysis, and initial value setting of Newton-Raphson method using bit pattern, a new regularity that can quickly encode and decode data into BL-beta code was discovered, and by applying this, the encoding speed of the algorithm was improved by an average of 24.8% and the decoding speed by an average of 5.3% compared to previous study.

• Korean Journal of Agricultural Science
• /
• v.49 no.3
• /
• pp.619-627
• /
• 2022

In Korea, apples and pears are among the most important agricultural products to farmers who seek to earn money as income. Generally, farmers make decisions at various stages to maximize their income but they do not always know exactly which option will be the best one. Many previous studies were conducted to solve this problem by predicting farmers' income structure, but researchers are still exploring better approaches. Currently, machine learning technology is gaining attention as one of the new approaches for farmers' income prediction. The machine learning technique is a methodology using an algorithm that can learn independently through data. As the level of computer science develops, the performance of machine learning techniques is also improving. The purpose of this study is to predict the income structure of apples and pears using the automatic machine learning solution H2O.AI and to present some implications for apple and pear farmers. The automatic machine learning solution H2O.AI can save time and effort compared to the conventional machine learning techniques such as scikit-learn, because it works automatically to find the best solution. As a result of this research, the following findings are obtained. First, apple farmers should increase their gross income to maximize their income, instead of reducing the cost of growing apples. In particular, apple farmers mainly have to increase production in order to obtain more gross income. As a second-best option, apple farmers should decrease labor and other costs. Second, pear farmers also should increase their gross income to maximize their income but they have to increase the price of pears rather than increasing the production of pears. As a second-best option, pear farmers can decrease labor and other costs.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.2
• /
• pp.381-404
• /
• 2022

Forest fires inflict great losses of human lives and serious damages to ecological systems. Hence, numerous fire detection methods have been proposed, one of which is fire detection based on sensors. However, these methods reveal several limitations when applied in large spaces like forests such as high cost, high level of false alarm, limited battery capacity, and other problems. In this research, we propose a novel forest fire detection method based on image processing and correlation coefficient. Firstly, two fire detection conditions are applied in RGB color space to distinguish between fire pixels and the background. Secondly, the image is converted from RGB to YCbCr color space with two fire detection conditions being applied in this color space. Finally, the correlation coefficient is used to distinguish between fires and objects with fire-like colors. Our proposed algorithm is tested and evaluated on eleven fire and non-fire videos collected from the internet and achieves up to 95.87% and 97.89% of F-score and accuracy respectively in performance evaluation.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.8
• /
• pp.39-48
• /
• 2023

This study was conducted to provide a solution to the problem of sizing errors occurring in online purchases due to discrepancies and non-standardization in clothing sizes. This paper discusses an implementation approach for a machine learning-based recommender system capable of providing personalized sizes to online consumers. We trained multiple validated collaborative filtering algorithms including Non-Negative Matrix Factorization (NMF), Singular Value Decomposition (SVD), k-Nearest Neighbors (KNN), and Co-Clustering using purchasing data derived from online commerce and compared their performance. As a result of the study, we were able to confirm that the NMF algorithm showed superior performance compared to other algorithms. Despite the characteristic of purchase data that includes multiple buyers using the same account, the proposed model demonstrated sufficient accuracy. The findings of this study are expected to contribute to reducing the return rate due to sizing errors and improving the customer experience on e-commerce platforms.

• International conference on construction engineering and project management
• /
• 2022.06a
• /
• pp.768-775
• /
• 2022

Current MEP designs are mostly applied by 2D-based design methods and tend to focus on simple modeling or geometry information expression such as converting 2D-written drawings into 3D modeling without taking advantage of the strength of BIM application. To increase the demand for BIM-based MEP design, geometric information, and property information of each member of the 3D model must be conveniently linked from the phase of the Design Development (DD) to the phase of Construction Document (CD). To conveniently implement a detailed model at each phase, the detailed level of each member of the 3D model must be specific, and an automatic generation of objects at each phase and automatic detailing module for each LOD are required. However, South Korea's guidelines have comprehensive standards for the degree of MEP modeling details for each design phase, and the application of each design phase is ambiguous. Furthermore, in practice, detailed levels of each phase are input manually. Therefore, this paper summarized the detailed standards of MEP modeling for each design phase through interviews with MEP design companies and related literature research. In addition, items that enable auto-detailing with DYNAMO were selected using the checklist for each design phase, and the types of detailed methods were presented. Auto-detailing items considering the detailed level of each phase were classified by members. If a DYNAMO algorithm is produced that automates selected auto-detailing items in this paper, the time and costs required for modeling construction will be reduced, and the demand for MEP design will increase.

• Atmosphere
• /
• v.33 no.1
• /
• pp.49-59
• /
• 2023

Blocking refers to a class of weather phenomena appearing in the mid and high latitudes, whose characteristics are blocked airflow of persistence. Frequently found over the Pacific and Atlantic regions of the Northern Hemisphere, blocking affects severe weather in the surrounding areas with different mechanisms depending on the type of blocking patterns. Along with lots of studies about persistent weather extremes focusing on the specific types of blocking, a new categorization using Rossby wave breaking has emerged. This study aims to apply this concept to the classification of blockings over the Pacific and examine how different wave breakings specify the associated cold weather in the Korean peninsula. At the same time, we investigate a strongly developing ridge around the Pacific by designing a new detection algorithm, where a reversal method is modified to distinguish ridge-type blocking patterns. As result, Kamchatka blocking (KB) and strong ridge over the Central Pacific are observed the most frequently during 20 years (2001~2020) of the studied period, and anomalous low pressures with cold air over the Korean Peninsula are accompanied by blocking events. When it considers the Rossby wave breaking, cyclonic wave-breaking is dominant in KB, which generates low-pressure anomalies over the Korean Peninsula. However, KB with anticyclone wave breaking appears with the high-pressure anomalies over the Korean Peninsula and it generates the warm temperature anomaly. Lastly, the low-pressure anomalies are also generated by the strong ridge over the Central Pacific, which persists for approximately three days and give a significant impact on cold surge on the Korean Peninsula.

• Journal of IKEEE
• /
• v.27 no.1
• /
• pp.1-11
• /
• 2023

Path driving tests are necessary for the development of self-driving cars or robots. These tests are being conducted in simulation as well as real environments. In particular, for development using reinforcement learning and deep learning, development through simulators is also being carried out when data of various environments are needed. To this end, it is necessary to utilize not only manually designed paths but also various randomly and automatically designed paths. This test site design can be used for actual construction and manufacturing. In this paper, we introduce a method for randomly generating a driving test path consisting of a combination of arcs and segments. This consists of a method of determining whether there is a collision by obtaining the distance between an arc and a line segment, and an algorithm that deletes part of the path and recreates an appropriate path if it is impossible to continue the path.