• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2006.10a
• /
• pp.176-183
• /
• 2006

A novel intelligent anti-collision decision-making supporting system is addressed in this paper. To obtain precise anti-collision information capability, an innovative neurofuzzy network is proposed and applied. A fuzzy set interpretation is incorporated into the network design to handle imprecise information. A neural network architecture is used to train the parameters of the Fuzzy Inference System (FIS). The learning process is based on a hybrid learning algorithm and off-line training data. The training data are obtained by trial manoeuvre. This neurofuzzy network can be considered to be a self-learning system with the ability to learn new information adaptively without forgetting old knowledge. This supporting system can decrease ship operators' burden to deal with bridge data and help them to make a precise anti-collision decision.

• Journal of Engineering Education Research
• /
• v.25 no.5
• /
• pp.55-67
• /
• 2022

Distance learning has become an efficient tool and is being widely used at work and in school. This research presents the results of a project-oriented entrepreneurship course taught both in classroom and online for a period of 3 years before and after the pandemic caused by COVID-19. Despite the various challenges, the outcome demonstrated that the students were able to attain required knowledge and capabilities in the online learning environment. As such, this paper discusses effective ways to blend in distance learning components so that both instructors and students can benefit from Internet-based education technologies. In the future, both face-to-face and virtual project work and study are likely to get integrated into a 'hyflex' class, which is flexible, on-offline education.

• Journal of information and communication convergence engineering
• /
• v.21 no.1
• /
• pp.9-16
• /
• 2023

The natural language on social network platforms has a certain front-to-back dependency in structure, and the direct conversion of Chinese text into a vector makes the dimensionality very high, thereby resulting in the low accuracy of existing text classification methods. To this end, this study establishes a deep learning model that combines a big data ultra-deep convolutional neural network (UDCNN) and long short-term memory network (LSTM). The deep structure of UDCNN is used to extract the features of text vector classification. The LSTM stores historical information to extract the context dependency of long texts, and word embedding is introduced to convert the text into low-dimensional vectors. Experiments are conducted on the social network platforms Sogou corpus and the University HowNet Chinese corpus. The research results show that compared with CNN + rand, LSTM, and other models, the neural network deep learning hybrid model can effectively improve the accuracy of text classification.

• Journal of Engineering Education Research
• /
• v.26 no.4
• /
• pp.36-46
• /
• 2023

This paper presents a case study and online-offline (hybrid) course structure for project-oriented community-based service learning in order to solve real-world problems for society. It examines social issues and conduct student projects to develop solutions that can generate sustainable value. This course helps students to use their major knowledge to assess and solve the problems faced by the local community. The outcomes of this course conducted via online lectures and offline project activities are discussed. The operation of this blended type of social problem-solving course is also stated.

• ETRI Journal
• /
• v.46 no.1
• /
• pp.137-153
• /
• 2024

This study introduces a new approach in Korean morphological analysis combining dictionary-based techniques with Transformer-based deep learning models. The key innovation is the use of a BERT-based reranking system, significantly enhancing the accuracy of traditional morphological analysis. The method generates multiple suboptimal paths, then employs BERT models for reranking, leveraging their advanced language comprehension. Results show remarkable performance improvements, with the first-stage reranking achieving over 20% improvement in error reduction rate compared with existing models. The second stage, using another BERT variant, further increases this improvement to over 30%. This indicates a significant leap in accuracy, validating the effectiveness of merging dictionary-based analysis with contemporary deep learning. The study suggests future exploration in refined integrations of dictionary and deep learning methods as well as using probabilistic models for enhanced morphological analysis. This hybrid approach sets a new benchmark in the field and offers insights for similar challenges in language processing applications.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.23 no.4
• /
• pp.110-126
• /
• 2024

Several studies have focused on developing the safest and most efficient path from the current location to the available parking area for vehicles entering a parking lot. In the present study, the parking lot structure and parking environment such as the lane width, width, and length of the parking space, were vaired by referring to the actual parking lot with vertical and horizontal parking. An automatic parking path planning model was proposed by collecting path data by various setting angles and environments such as a starting point and an arrival point, by putting the collected data into a deep learning model. The existing algorithm(Hybrid A-star, Reeds-Shepp Curve) and the deep learning model generate similar paths without colliding with obstacles. The distance and the consumption time were reduced by 0.59% and 0.61%, respectively, resulting in more efficient paths. The switching point could be decreased from 1.3 to 1.2 to reduce driver fatigue by maximizing straight and backward movement. Finally, the path generation time is reduced by 42.76%, enabling efficient and rapid path generation, which can be used to create a path plan for autonomous parking during autonomous driving in the future, and it is expected to be used to create a path for parking robots that move according to vehicle construction.

• Korean Journal of Clinical Pharmacy
• /
• v.32 no.4
• /
• pp.281-287
• /
• 2022

Pharmacists should maintain professional competencies to provide optimal pharmaceutical care services to patients, which can be achieved through continued commitment to lifelong learning. Traditionally continuing education (CE) has been widely used as a way of lifelong learning for many healthcare professionals. It, however, has several limitations. CE is delivered in the form of instructor-led education focused on multiple learners. Learning is passive and reactive for participants, so it sometimes does not lead to bringing behavioral changes in workplace performance. Therefore, recently the concept of lifelong learning tends to move from CE toward continuing professional development (CPD). CPD is an ongoing process that improves knowledge, skills, and competencies throughout a professional's career. It is a more comprehensive structured approach toward the enhancement of personal competencies. It emphasizes an individual's learning needs and goals and enables learning to become proactive, conscious, and self-directed. CPD consists of four stages: reflect, plan, learn, and evaluate. CE is one component of CPD. Each stage is recorded in a CPD portfolio. There are many practical difficulties in implementing the complete CPD system for lifelong learning of pharmacists in many countries including Korea. Applying a hybrid form that utilizes CPD and CE together, as in the case of some countries, could be an alternative. Furthermore, in undergraduate pharmacy education, it is necessary to teach students about CPD and train them on how to perform CPD as a pharmacist.

• Computers and Concrete
• /
• v.32 no.3
• /
• pp.263-286
• /
• 2023

Experimenting with concrete to determine its compressive and tensile strengths is a laborious and time-consuming operation that requires a lot of attention to detail. Researchers from all around the world have spent the better part of the last several decades attempting to use machine learning algorithms to make accurate predictions about the technical qualities of various kinds of concrete. The research that is currently available on estimating the strength of concrete draws attention to the applicability and precision of the various machine learning techniques. This article provides a summary of the research that has previously been conducted on estimating the strength of concrete by making use of a variety of different machine learning methods. In this work, a classification of the existing body of research literature is presented, with the classification being based on the machine learning technique used by the researchers. The present review work will open the horizon for the researchers working on the machine learning based prediction of the compressive strength of concrete by providing the recommendations and benefits and drawbacks associated with each model as determining the compressive strength of concrete practically is a laborious and time-consuming task.

• The Journal of the Korea Contents Association
• /
• v.18 no.7
• /
• pp.495-502
• /
• 2018

The advent of deep learning technology has made rapid progress in handwritten letter recognition in many languages. Handwritten Chinese recognition has improved to 97.2% accuracy while handwritten Japanese recognition approached 99.53% percent accuracy. Hanguel handwritten letters have many similar characters due to the characteristics of Hangeul, so it was difficult to recognize the letters because the number of data was small. In the handwritten Hanguel recognition using Hybrid Learning, it used a low layer model based on lenet and showed 96.34% accuracy in handwritten Hanguel database PE92. In this paper, 98.64% accuracy was obtained by organizing deep CNN (Convolution Neural Network) in handwritten Hangeul recognition. We designed a new network for handwritten Hangeul data based on GoogLenet without using the data augmentation or the multitasking techniques used in Hybrid learning.

• The KIPS Transactions:PartC
• /
• v.10C no.6
• /
• pp.705-710
• /
• 2003

Recently, most interchanges of information have been performed in the internet environments. So, the technuque, which is used as intrusion deleting tool for system protecting against attack, is very important. But, the skills of intrusion detection are newer and more delicate, we need preparations for defending from these attacks. Currently, lots of intrusion detection systemsmake the midel of intrusion detection rule using experienced data, based on this model they have the strategy of defence against attacks. This is not efficient for defense from new attack. In this paper, a new model of intrusion detection is proposed. This is hybrid statistical learning model using likelihood ratio test and statistical learning theory, then this model can detect a new attack as well as experienced attacks. This strategy performs intrusion detection according to make a model by finding abnomal attacks. Using KDD Cup-99 task data, we can know that the proposed model has a good result of intrusion detection.