• 디지털산업정보학회논문지
• /
• 제19권3호
• /
• pp.233-244
• /
• 2023

Although artificial intelligence and machine learning technologies have been used in various fields, problems with personal information protection have arisen based on centralized data collection and processing. Federated learning has been proposed to solve this problem. Federated learning is a process in which clients who own data in a distributed data environment learn a model using their own data and collectively create an artificial intelligence model by centrally collecting learning results. Unlike the centralized method, Federated learning has the advantage of not having to send the client's data to the central server. In this paper, we quantitatively present the performance improvement when federated learning is applied using the building change detection learning data. As a result, it has been confirmed that the performance when federated learning was applied was about 29% higher on average than the performance when it was not applied. As a future work, we plan to propose a method that can effectively reduce the number of federated learning rounds to improve the convergence time of federated learning.

• International Journal of Advanced Culture Technology
• /
• 제5권2호
• /
• pp.74-81
• /
• 2017

Recently, there are many companies that use systems based on artificial intelligence. The accuracy of artificial intelligence depends on the amount of learning data and the appropriate algorithm. However, it is not easy to obtain learning data with a large number of entity. Less data set have large generalization errors due to overfitting. In order to minimize this generalization error, this study proposed DGA(Dropout Genetic Algorithm) which can expect relatively high accuracy even though data with a less data set is applied to machine learning based genetic algorithm to deep learning based dropout. The idea of this paper is to determine the active state of the nodes. Using Gradient about loss function, A new fitness function is defined. Proposed Algorithm DGA is supplementing stochastic inconsistency about Dropout. Also DGA solved problem by the complexity of the fitness function and expression range of the model about Genetic Algorithm As a result of experiments using MNIST data proposed algorithm accuracy is 75.3%. Using only Dropout algorithm accuracy is 41.4%. It is shown that DGA is better than using only dropout.

• 한국멀티미디어학회논문지
• /
• 제25권2호
• /
• pp.311-324
• /
• 2022

In this paper, we adjusted the feedback and learning materials for each learning based on ARCS motivation which applied DEVS methodology. We designed the ARCS professor-student model that expresses the continuous change in the student's attitude toward the class according to the student's attention, relevance, confidence, and satisfaction. It was applied to computational thinking and data analysis classes Based on the designed model. Before and after class, the students were asked the same question and then analyzed for each part of the ARCS. It was observed that students' perceptions of Attention, Relevance, and Satisfaction were improved except for Confidence. we observed that the students themselves felt that they lacked a lot of confidence compared to other ARS through the analysis. Although, Confidence showed a 13.5% improvement after class but it was about 33% lower than the average of other ARS. However, when it was observed that students' self-confidence was 30% lower than other motivational factors it was confirmed that the part that leads C to a similar level in other ARS is necessary.

• 한국임상약학회지
• /
• 제34권2호
• /
• pp.134-139
• /
• 2024

Background: Increasing numbers of studies and research about artificial intelligence (AI) and machine learning (ML) have led to their application in clinical trials. The purpose of this study is to analyze computer-based new technologies (AI/ML) applied on clinical trials registered on ClinicalTrials.gov to elucidate current usage of these technologies. Methods: As of March 1st, 2023, protocols listed on ClinicalTrials.gov that claimed to use AI/ML and included at least one of the following interventions-Drug, Biological, Dietary Supplement, or Combination Product-were selected. The selected protocols were classified according to their context of use: 1) drug discovery; 2) toxicity prediction; 3) enrichment; 4) risk stratification/management; 5) dose selection/optimization; 6) adherence; 7) synthetic control; 8) endpoint assessment; 9) postmarketing surveillance; and 10) drug selection. Results: The applications of AI/ML were explored in 131 clinical trial protocols. The areas where AI/ML was most frequently utilized in clinical trials included endpoint assessment (n=80), followed by dose selection/optimization (n=15), risk stratification/management (n=13), drug discovery (n=4), adherence (n=4), drug selection (n=1) and enrichment (n=1). Conclusion: The most frequent application of AI/ML in clinical trials is in the fields of endpoint assessment, where the utilization is primarily focuses on the diagnosis of disease by imaging or video analyses. The number of clinical trials using artificial intelligence will increase as the technology continues to develop rapidly, making it necessary for regulatory associates to establish proper regulations for these clinical trials.

• 한국환경복원기술학회지
• /
• 제24권6호
• /
• pp.89-96
• /
• 2021

In this study, a deep-learning image analysis model was established and validated for AI-based monitoring of the tidal flat ecosystem for marine protected creatures Ocypode stimpsoni and their habitat. The data in the study was constructed using an unmanned aerial vehicle, and the U-net model was applied for the deep learning model. The accuracy of deep learning model learning results was about 0.76 and about 0.8 each for the Ocypode stimpsoni and their burrow whose accuracy was higher. Analyzing the distribution of crabs and burrows by putting orthomosaic images of the entire study area to the learned deep learning model, it was confirmed that 1,943 Ocypode stimpsoni and 2,807 burrow were distributed in the study area. Through this study, the possibility of using the deep learning image analysis technology for monitoring the tidal ecosystem was confirmed. And it is expected that it can be used in the tidal ecosystem monitoring field by expanding the monitoring sites and target species in the future.

• 인터넷정보학회논문지
• /
• 제24권6호
• /
• pp.119-125
• /
• 2023

우크라이나-러시아 전을 통해 드론의 군사적 가치는 재평가되고 있으며, 북한은 '22년 말 대남 드론 도발을 통해 실제 검증까지 완료한 바 있다. 또한, 북한은 인공지능(AI) 기술의 드론 적용을 추진하고 있는 것으로 드러나 드론의 위협은 나날이 커지고 있다. 이에 우리 군은 드론작전사령부를 창설하고 다양한 드론 대응 체계를 도입하는 등 대 드론 체계 구축을 도모하고 있지만, 전력증강 노력이 타격체계 위주로 편중되어 군집드론 공격에 대한 효과적 대응이 우려된다. 특히, 도심에 인접한 공군 비행단은 민간 피해가 우려되어 재래식 방공무기의 사용 역시 극도로 제한되는 실정이다. 이에 본 연구에서는 AI기술이 적용된 적 군집드론의 위협으로부터 아 항공기의 생존성 향상을 위해 AI모델의 객체탐지 능력을 저해하는 소극방공 기법을 제안한다. 대표적인 적대적 머신러닝(Adversarial machine learning) 기술 중 하나인 적대적 예제(Adversarial example)를 레이저를 활용하여 항공기에 조사함으로써, 적 드론에 탑재된 객체인식 AI의 인식률 저하를 도모한다. 합성 이미지와 정밀 축소모형을 활용한 실험을 수행한 결과, 제안기법 적용 전 약 95%의 인식률을 보이는 객체인식 AI의 인식률을 제안기법 적용 후 0~15% 내외로 저하시키는 것을 확인하여 제안기법의 실효성을 검증하였다.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2021년도 추계학술대회
• /
• pp.385-387
• /
• 2021

해양 침적 쓰레기는 저서 생태계를 위협하고 유령어업으로 인한 어획량 감소를 초래하여 연간 약 3,700억 원의 피해를 발생시키고 있다. 이를 수거하기 위해서 현재 양방향 음파탐지기와 잠수, 인양틀 등을 이용하여 현황조사를 수행한다. 하지만, 많은 침적 쓰레기를 조사하기엔 조사범위가 적고 인명피해를 불러올 가능성이 있다. 본 논문에서는 실시간 객체 탐지에 적합한 YOLOv5 알고리즘을 활용하여 AI-Hub의 해안 침적 쓰레기 이미지 데이터를 학습시켜 높은 정확도의 해양 침적 쓰레기 감지 인공지능 모델을 구현한 내용을 다룬다.

• International Journal of Computer Science & Network Security
• /
• 제23권12호
• /
• pp.225-234
• /
• 2023

In this era of information and communication technology (ICT), tremendous improvements have been witnessed in our daily lives. The impact of these technologies is subjective and negative or positive. For instance, ICT has brought a lot of ease and versatility in our lifestyles, on the other hand, its excessive use brings around issues related to physical and mental health etc. In this study, we are bridging these both aspects by proposing the idea of AI based mental healthcare (AIMS). In this regard, we aim to provide a platform where the patient can register to the system and take consultancy by providing their assessment by means of a chatbot. The chatbot will send the gathered information to the machine learning block. The machine learning model is already trained and predicts whether the patient needs a treatment by classifying him/her based on the assessment. This information is provided to the mental health practitioner (doctor, psychologist, psychiatrist, or therapist) as clinical decision support. Eventually, the practitioner will provide his/her suggestions to the patient via the proposed system. Additionally, the proposed system prioritizes care, support, privacy, and patient autonomy, all while using a friendly chatbot interface. By using technology like natural language processing and machine learning, the system can predict a patient's condition and recommend the right professional for further help, including in-person appointments if necessary. This not only raises awareness about mental health but also makes it easier for patients to start therapy.

• 융합정보논문지
• /
• 제12권5호
• /
• pp.139-149
• /
• 2022

본 연구는 예비교사의 미래 교육을 위한 인공지능 융합교육 역량을 함양하고, 동시에 학생의 학습 과정에 대한 이해를 증진할 수 있는 교육 프로그램을 개발하고 효과를 분석하는 것을 목적으로 하였다. 이를 위해 물질의 상태 분류를 주제로 머신러닝포키즈와 스크래치3를 활용한 인공지능 융합교육 프로그램을 15주차 분량으로 개발하였다. 개발된 내용은 자발적으로 참여한 K대학교 예비교사들에게 처치되었다. 그 결과, 예비교사들은 머신러닝의 학습을 이해하는 과정을 통해 학생의 학습 과정을 비유적으로 이해할 수 있었다. 또한, 인공지능 교수효능감의 사전-사후 t검정 결과는 t=-7.137(p< .000)으로 통계적으로 유의한 향상을 보였다. 따라서 본 연구에서 개발한 인공지능 융합교육 프로그램은 교생실습 외에 비간접적인 방식으로 예비교사의 학생에 대한 이해를 높일 수 있는데 도움이 되고, 인공지능 교육 역량 함양에 기여할 수 있음이 시사된다.

• 한국콘텐츠학회논문지
• /
• 제22권4호
• /
• pp.106-117
• /
• 2022

2021년에는 고등학교 교육과정에서 인공지능기초 과목이 도입되었다. 과거 교육에 ICT 접목 시 발생했던 원리 보다 활용 중심 교육의 문제가 인공지능 교육의 도입에도 되풀이될까 우려하는 의견이 다수 존재한다. 기존의 인공지능 교육 플랫폼은 대부분 인공지능의 활용에만 초점을 두고 있다. 중학생의 인공지능 교육을 위해 인공지능이 결과를 도출하는 과정에 관한 학습과 인공지능 알고리즘의 원리 학습에 어려움이 있다. 최근 메타버스의 교육적 적용이 화두가 되면서, 학생들의 몰입감과 흥미를 유발해 학습성취도 향상에 도움을 주려는 연구들이 시작되었다. 본 연구는 중학교 인공지능 교육에서 활용될 수 있는 교육 콘텐츠로 메타버스를 활용한 의사결정나무 알고리즘에 관한 교육 게임 콘텐츠를 개발하였다. 게임을 교육에 접목함으로써 학생들의 인공지능에 대해 흥미와 몰입감을 높이고 교육 효과를 높이고자 하였다. 본 연구에서는 개발한 게임 콘텐츠에 관하여 예비 교원들을 대상으로 교육 효과성, 난이도, 흥미 정도를 조사·분석하였고, 이를 바탕으로 향후 원리 중심의 인공지능 교육 방법에 관하여 제언하였다.