• Electronics and Telecommunications Trends
• /
• v.35 no.4
• /
• pp.81-90
• /
• 2020

The use of machine learning technologies such as deep and reinforcement learning has proliferated in various domains with the advancement of deep neural network studies. To make the learning successful, both big data acquisition and fast processing are required. However, for some physical world applications such as autonomous drone flight, it is difficult to achieve efficient learning because learning with a premature A.I. is dangerous, cost-ineffective, and time-consuming. To solve these problems, simulation-based approaches can be considered. In this study, we analyze recent trends in drone simulation technologies and compare their features. Subsequently, we introduce Octopus, which is a highly precise and scalable drone simulator being developed by ETRI.

• Journal of IKEEE
• /
• v.24 no.2
• /
• pp.514-522
• /
• 2020

In Continuous Learning environment, we identify that the Catastrophic Forgetting phenomenon, which forgets the information of previously learned data, occurs easily between data having different domains. To control this phenomenon, we introduce how to measure the relationship between previously learned data and newly learned data through the distribution of the neural network's output, and how to use these measurements to mitigate the Catastrophic Forcing phenomenon. MNIST and EMNIST data were used for evaluation, and experiments showed an average 22.37% improvement in accuracy for previous data.

• The Journal of Korea Robotics Society
• /
• v.10 no.2
• /
• pp.90-95
• /
• 2015

The main goal of e-learning systems is just-in-time knowledge acquisition. Rule-based e-learning systems, however, suffer from the mesa effect and the cold start problem, which both result in low user acceptance. E-learning systems suffer a further drawback in rendering the implementation of a natural interface in humanoids difficult. To address these concerns, even exceptional questions of the learner must be answerable. This paper aims to propose a method that can understand the learner's verbal cues and then intelligently explore additional domains of knowledge based on crowd data sources such as Wikipedia and social media, ultimately allowing for better answers in real-time. A prototype system was implemented using the NAO platform.

• Journal of Information Processing Systems
• /
• v.15 no.2
• /
• pp.320-330
• /
• 2019

This study analyzed changes in sociality and democratic-citizenship among elementary school students in the information class and the science class at the Science Education Institute for the Gifted, who were divided into an experimental group and a control group. The experimental group engaged in the Learning Together (LT) cooperative form of learning for which the remix function of Scratch, an educational programming language, was applied, while the control group was given general instructor-led lessons. Members in the experimental group were able to modify processes during projects through the usage of the remix function, thereby actively participating in the projects and eventually generating team-based results. The post-class t-tests showed a greater degree of improvements in sociality and democratic citizenship for the experimental group that was offered the remix-function-based cooperative learning than the control group. Statistically significant differences were present between two groups particularly in "cooperative spirit" sub-domain of sociality and the "community" and "responsibility" sub-domains of democratic citizenship.

• 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.

• ETRI Journal
• /
• v.44 no.1
• /
• pp.155-167
• /
• 2022

The Internet of Things (IoT) is a new paradigm that connects physical and virtual objects from various domains such as home automation, industrial processes, human health, and monitoring. IoT sensors receive information from their environment and forward it to their neighboring nodes. However, the large amounts of exchanged data are vulnerable to attacks that reduce the network performance. Most of the previous security methods for IoT have neglected the energy consumption of IoT, thereby affecting the performance and reducing the network lifetime. This paper presents a new multistep routing protocol based on cellular learning automata. The network lifetime is improved by a performance-based adaptive reward and fine parameters. Nodes can vote on the reliability of their neighbors, achieving network reliability and a reasonable level of security. Overall, the proposed method balances the security and reliability with the energy consumption of the network.

• Journal of Korean society of Dental Hygiene
• /
• v.21 no.2
• /
• pp.193-201
• /
• 2021

Objectives: We evaluated the learning objectives of dental hygiene courses based on Bloom's learning objectives, and analyze the degree of match with the dental hygienist's job for each detailed subject. Methods: The 5th edition of 'Dental hygiene and learning objectives' was analyzed by subject based on Bloom's cognitive domain classification from March 10 to April. In addition, the degree of match between the contents of the secondary job analysis of the dental hygienist and the learning objectives for each detailed subject were analyzed. Results: The total number of dental hygiene learning objectives was 2,975 (2,762 theory, 52 practice). Among the cognitive domains, the comprehension domain was the most common (79.8%), and the skill domain was very low (4.9%). In the job for each detailed subject of dental hygiene, the most frequently performed was 'dental prophylaxis and practice' with 103 subjects. Conclusions: Overall, dental hygiene learning objectives are mostly theory-oriented, so it is necessary to expand and improve in the direction related to the jobs that clinical dental hygienists perform in the field. In addition, it is necessary to continuously develop timely learning goals, and prepare active strategies for developing high-quality items.

• Journal of Internet Computing and Services
• /
• v.24 no.4
• /
• pp.37-45
• /
• 2023

Adaptive has gained significant attention in Education Technology (EdTech), with personalized learning experiences becoming increasingly important. Next-generation chatbots, including models like ChatGPT, are emerging in the field of education. These advanced tools show great potential for delivering personalized and adaptive learning experiences. This paper reviews previous research on adaptive learning and the role of chatbots in education. Based on this, the paper explores current and future chatbot technologies to propose a framework for using ChatGPT or similar chatbots in adaptive learning. The framework includes personalized design, targeted resources and feedback, multi-turn dialogue models, reinforcement learning, and fine-tuning. The proposed framework also considers learning attributes such as age, gender, cognitive ability, prior knowledge, pacing, level of questions, interaction strategies, and learner control. However, the proposed framework has yet to be evaluated for its usability or effectiveness in practice, and the applicability of the framework may vary depending on the specific field of study. Through proposing this framework, we hope to encourage learners to more actively leverage current technologies, and likewise, inspire educators to integrate these technologies more proactively into their curricula. Future research should evaluate the proposed framework through actual implementation and explore how it can be adapted to different domains of study to provide a more comprehensive understanding of its potential applications in adaptive learning.

• The Journal of the Korea Contents Association
• /
• v.22 no.10
• /
• pp.437-444
• /
• 2022

The purpose of this study was to investigate the effects of Read Along by Google with primary English learners' pronunciation and affective domains. In order to answer these two questions, a 4-week pilot study was conducted with 24 participants in the 6 th grade. Read Along as a main learning tool was utilized for a reading-aloud activity, and a pre-/post pronunciation test and survey on the affective factors were distributed as a research instrument. The results indicated that a read-aloud activity with Read Along brought a positive impact on the development of learners' pronunciation ability in terms of accuracy and fluency. Participants showed improvement in the post-pronunciation test, compared to the pre-one and there was a significant difference based on the result of the paired samples t-test. Next, the results of the survey on the affective domains illustrated that participants showed overall improvement in learning interest and confidence and there was a significant difference in these factors. Yet, there was not a significant difference in the learning attitude, even though they showed partial improvement.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.18 no.6
• /
• pp.187-196
• /
• 2017

In using lecture evaluation methods to improve the quality of education, most universities need to reflect the changes in the educational environment. The transformation of university education into a mixed learning environment blending face-to-face education and online education necessitates the development of appropriate lecture evaluation items. For this purpose, we analyzed the items and the factor analysis for the students of C university in Daejeon. The primary data were carried out with 47 measurement items in 10 domains, such as planning and preparation of lectures, which were found through previous research analysis. Secondary data were validated by using the items confirmed through analysis of preliminary test data. The study results showed that 20 items including six domains such as planning and preparation of lectures, learning materials, learning tasks, instruction media, online course test and grades were derived. These study results suggest that universities lectures should be evaluated to ensure improvement.