• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.5
• /
• pp.75-86
• /
• 2022

This study aims to solve the problem of class imbalance in numerical data by using a deep learning-based Variational AutoEncoder and to improve the performance of the learning model by augmenting the learning data. We propose 'D-VAE' to artificially increase the number of records for a given table data. The main features of the proposed technique go through discretization and feature selection in the preprocessing process to optimize the data. In the discretization process, K-means are applied and grouped, and then converted into one-hot vectors by one-hot encoding technique. Subsequently, for memory efficiency, sample data are generated with Variational AutoEncoder using only features that help predict with RFECV among feature selection techniques. To verify the performance of the proposed model, we demonstrate its validity by conducting experiments by data augmentation ratio.

• Journal of Korea Trade
• /
• v.25 no.2
• /
• pp.58-74
• /
• 2021

Purpose - This paper research on the embodied carbon emission in Sino-Korea trade. It calculates and analyzes the carbon emission coefficient and specific carbon emissions in Sino-Korea trade from 2005 to 2014. Design/methodology - This paper conducted an empirical analysis for embodied carbon emission in Sino-Korea trade during the years 2005-2014, using a multi-region input-output model. First, direct and complete CO2 emission coefficient of the two countries were calculated and compared. On this basis, combined with the world input-output table, the annual import and export volume and sector volume of embodied carbon emission are determined. Then through the comparative analysis of the empirical results, the reasons for the carbon imbalance in Sino-Korea trade are clarified, and the corresponding suggestions are put forward according to the environmental protection policies being implemented by the two countries. Findings - The results show that South Korea is in the state of net trade export and net embodied carbon import. The carbon emission coefficient of most sectors in South Korea is lower than that of China. However, the reduction of carbon emission coefficient in China is significantly faster than that in South Korea in this decade. The change of Korea's complete CO2 emission coefficient shows that policy factors have a great impact on environmental protection. The proportion of intra industry trade between China and South Korea is relatively large and concentrated in mechanical and electrical products, chemical products, etc. These sectors generally have large carbon emissions, which need to be noticed by both countries. Originality/value - To the best knowledge of the authors, this study is the first attempt to research the embodied carbon emission of ten consecutive years in Sino-Korea Trade. In addition, In this paper, some mathematical methods are used to overcome the error problem caused by different statistical caliber in different databases. Finally, the accurate measurement of carbon level in bilateral trade will provide some reference for trade development and environmental protection.

• Journal of Digital Policy
• /
• v.1 no.2
• /
• pp.17-24
• /
• 2022

In China, the continuous promotion and coverage of scanning code payment has caused an imbalance in the coin market. Coin changers can not only alleviate this problem, but also affect the business development of commercial banks. Therefore, it is meaningful to study the coin changer. The purpose of this paper is to study the impact of coin changers on the business of commercial banks in China. Through on-the-spot visits and based on the manually collected customer data of Chinese commercial banks as the object, combined with the calculation method of financial indicators to conduct case analysis. The results of the study show that the coin changer has a positive impact on the business development of Chinese commercial banks. This paper provides feasible suggestions and new ideas for business development to Chinese commercial banks. At present, there are few related studies on coin exchange machines. This study combines the calculation of financial indicators to verify the policy results, which is the innovation of this study.

• Journal of Adhesion and Interface
• /
• v.23 no.2
• /
• pp.17-24
• /
• 2022

Colloidal quantum dots (QDs) have gained attention for applications in quantum dot light emitting diodes (QLEDs) due to their high photoluminescence quantum yield, narrow emission spectra, and tunable bandgap. Nevertheless, non-radiative recombination induced by electron and hole imbalance deteriorates the device efficiency and stability. To overcome the problem, researchers have been trying to enhance hole transport properties of hole transporting layers (HTL) and/or slow down the electron injection in electron transport layer (ETL). Here, we summarize two approaches: i) development of interfacial materials between QD and ETL (or HTL); ii) engineering of HTL by blending or multi-layer approaches.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.32 no.2
• /
• pp.201-211
• /
• 2022

According to the recent change in the cybersecurity paradigm, research on anomaly detection methods using machine learning and deep learning techniques, which are AI implementation technologies, is increasing. In this study, a comparative study on data preprocessing techniques that can improve the anomaly detection performance of a GRU (Gated Recurrent Unit) neural network-based intrusion detection model using NGIDS-DS (Next Generation IDS Dataset), an open dataset, was conducted. In addition, in order to solve the class imbalance problem according to the ratio of normal data and attack data, the detection performance according to the oversampling ratio was compared and analyzed using the oversampling technique applied with DCGAN (Deep Convolutional Generative Adversarial Networks). As a result of the experiment, the method preprocessed using the Doc2Vec algorithm for system call feature and process execution path feature showed good performance, and in the case of oversampling performance, when DCGAN was used, improved detection performance was shown.

• Journal of Korean Physical Therapy Science
• /
• v.29 no.4
• /
• pp.27-42
• /
• 2022

Background: Although children with cerebral palsy (CP) are able to walk independently, gait imbalance occurs due to abnormal muscle tone, musculoskeletal deformity, loss of balance, and selective motor control impairment. Gait restriction in the community and school is a major problem of rehabilitation in CP. Home-based therapy (HBT) provides a variety of interventions in which the therapist and the parent work together to resolve the activities and problems caused by the child's body structure. Therefore, we investigate the effectiveness of home-centered therapy on gross motor function in CP and try to present the possibility of clinical application. Design: A Systematic Review Methods: Research papers were published from Jan, 2012 to Jan, 2022 and were searched using Medline and PubMed. The search terms are 'family-centered' OR 'home-based' AND 'cerebral palsy'. A total of nine papers were analyzed in this study. The paper presented the quality level based on Physiotherapy Evidence Database (PEDro) scores to assess the quality of randomized clinical trials studies. Results: The results showed that HBT for strengthening exercise in lower extremity has a positive effect on the isokinetic torque and gross motor function. home-based treadmill therapy in CP is effective to perform at least 12 sessions of treadmill HBP in which the therapist determines the treadmill speed every week and the child's own gait pattern is modified. Conclusion: These results suggest that it will be important data for founding evidence on the effectiveness of home-centered therapy on gross motor function in children with cerebral palsy to advance clinical protocols.

• Smart Structures and Systems
• /
• v.31 no.4
• /
• pp.325-334
• /
• 2023

Computer vision-based damage detection enables non-contact, efficient and low-cost bridge health monitoring, which reduces the need for labor-intensive manual inspection or that for a large number of on-site sensing instruments. By leveraging recent semantic segmentation approaches, we can detect regions of critical structural components and identify damages at pixel level on images. However, existing methods perform poorly when detecting small and thin damages (e.g., cracks); the problem is exacerbated by imbalanced samples. To this end, we incorporate domain knowledge to introduce a hierarchical semantic segmentation framework that imposes a hierarchical semantic relationship between component categories and damage types. For instance, certain types of concrete cracks are only present on bridge columns, and therefore the noncolumn region may be masked out when detecting such damages. In this way, the damage detection model focuses on extracting features from relevant structural components and avoid those from irrelevant regions. We also utilize multi-scale augmentation to preserve contextual information of each image, without losing the ability to handle small and/or thin damages. In addition, our framework employs an importance sampling, where images with rare components are sampled more often, to address sample imbalance. We evaluated our framework on a public synthetic dataset that consists of 2,000 railway bridges. Our framework achieves a 0.836 mean intersection over union (IoU) for structural component segmentation and a 0.483 mean IoU for damage segmentation. Our results have in total 5% and 18% improvements for the structural component segmentation and damage segmentation tasks, respectively, compared to the best-performing baseline model.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.8
• /
• pp.365-370
• /
• 2023

This paper aims to compare the performance of speech data classification into two groups, adult and elderly, based on the acoustic and linguistic characteristics that change due to aging, such as changes in respiratory patterns, phonation, pitch, frequency, and language expression ability. For acoustic features we used attributes related to the frequency, amplitude, and spectrum of speech voices. As for linguistic features, we extracted hidden state vector representations containing contextual information from the transcription of speech utterances using KoBERT, a Korean pre-trained language model that has shown excellent performance in natural language processing tasks. The classification performance of each model trained based on acoustic and linguistic features was evaluated, and the F1 scores of each model for the two classes, adult and elderly, were examined after address the class imbalance problem by down-sampling. The experimental results showed that using linguistic features provided better performance for classifying adult and elderly than using acoustic features, and even when the class proportions were equal, the classification performance for adult was higher than that for elderly.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.22 no.3
• /
• pp.51-61
• /
• 2023

Travel demand estimation of E-Scooter is the start point of solving the regional demand-supply imbalance problem and plays pivotal role in a linked transportation system such as Mobility-as-a-Service (a.k.a. MaaS). Most focuses on developing trip generation model of shared E-Scooter but it is no study on selection of an appropriate zone scale when it comes to estimating travel demand of E-Scooter. This paper aimed for selecting an optimal TAZ scale for developing trip distribution model for shared E-Scooter. The TAZ scale candidates were selected in 250m, 500m, 750m, 1,000m square grid. The shared E-Scooter usage historical data were utilized for calculating trip distance and time, and then applying to developing gravity model. Mean Squared Error (MSE) is applied for the verification step to select the best suitable gravity model by TAZ scale. As a result, 250m of TAZ scale is the best for describing practical trip distribution of shared E-Scooter among the candidates.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.161-163
• /
• 2021

This urban centralization is expected to develop rapidly, with 75% of the population living in the city by 2035. Large cities are becoming unsustainable due to side effects such as environmental pollution, severe traffic jams, excessive energy depletion, and destruction of the natural ecosystem. In addition, the happiness index of citizens of large cities is also falling because of high crime rates and safety accidents, the work-life imbalance caused by inequality and polarization, and overly competitive education. To solve this problem, Smart City, an IT-based future city model, was born. The Korean government is also actively attempting to improve urban competitiveness and promote sustainable development through efficient construction and operation of smart cities as a national focus project. To support the effort, we review the basic directions and strategies of Sejong City's Smart City service infrastructure based on the comprehensive national land plan, Smart City plan, and Smart City strategy plan.