• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.165-165
• /
• 2023

Streamflow prediction is a critical task in water resources management and essential for planning and decision-making purposes. However, the streamflow prediction is challenging due to the complexity and non-linear nature of hydrological processes. The transfer learning is a powerful technique that enables a model to transfer knowledge from a source domain to a target domain, improving model performance with limited data in the target domain. In this study, we apply the transfer learning using the Informer model, which is a state-of-the-art deep learning model for streamflow prediction. The model was trained on a large-scale hydrological dataset in the source basin and then fine-tuned using a smaller dataset available in the target basin to predict the streamflow in the target basin. The results demonstrate that transfer learning using the Informer model significantly outperforms the traditional machine learning models and even other deep learning models for streamflow prediction, especially when the target domain has limited data. Moreover, the results indicate the effectiveness of streamflow prediction when knowledge transfer is used to improve the generalizability of hydrologic models in data-sparse regions.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.12
• /
• pp.1138-1144
• /
• 2013

This paper proposes a tactile transfer and display method between a data glove and vibration motors module. The data glove is developed to capture the hand postures and to measure the grip forces. The measured data are simplified with the proposed 5-bit transfer and display algorithm, and the vibration motors module is developed to display the measured hand posture and grip force to the operator. The proposed 5-bit algorithm contains both an 8-step hand posture and 4-step grip force level information for tactile transfer to the vibration motors module. Also, the effectiveness of the proposed method is shown through several experiments.

• Proceedings of the KSME Conference
• /
• 2007.05b
• /
• pp.3395-3400
• /
• 2007

Heat transfer experiments at a vertical annulus passage were carried out in the SPHINX(Supercritical Pressure Heat Transfer Investigation for NeXt Generation) to investigate the heat transfer behaviors of supercritical $CO_2$. The collected test data are to be used for the reactor core design of the SCWR (SuperCritical Water-cooled Reactor). The mass flux was in the range of 400${\sim}$1200 kg/$m^2$s and the heat flux was chosen up to 150 kW/$m^2$. The selected pressures were 7.75 and 8.12 MPa. The heat transfer data were analyzed and compared with the previous tube test data. The test results showed that the heat transfer characteristics were similar to those of the tube in case of a normal heat transfer mode and degree of heat transfer deterioration became smaller than that in the tube. Comparison of the experimental heat transfer coefficients with the predicted ones by the existing correlations showed that there was not a distinct difference between the correlations.

• Proceedings of the IEEK Conference
• /
• 1998.10a
• /
• pp.515-518
• /
• 1998

This paper describes a direct transfer method of multimedia data stream between multimedia processor and network device without using system memory. The hardware architecture and functions for direct transfer, the method to transfer multimedia data to and from the multimedia processor and etc are described. Comparing the proposed method with general methods, I show that the direct transfer method can decrease number of bus accesses and bus cycles.

• Journal of Computing Science and Engineering
• /
• v.5 no.3
• /
• pp.257-268
• /
• 2011

Machine learning and data mining have found many applications in biological domains, where we look to build predictive models based on labeled training data. However, in practice, high quality labeled data is scarce, and to label new data incurs high costs. Transfer and multitask learning offer an attractive alternative, by allowing useful knowledge to be extracted and transferred from data in auxiliary domains helps counter the lack of data problem in the target domain. In this article, we survey recent advances in transfer and multitask learning for bioinformatics applications. In particular, we survey several key bioinformatics application areas, including sequence classification, gene expression data analysis, biological network reconstruction and biomedical applications.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.30 no.5 s.248
• /
• pp.389-396
• /
• 2006

For the successful design of heat exchangers, it is very important to understand local heat transfer phenomena on the circular tube of heat exchangers. In the present study, experiments are performed for single circular tube and tube banks with and without heat transfer promoters. The naphthalene sublimation technique is employed to measure the local mass transfer coefficients, and the measured local mass transfer data are converted to the local heat transfer data using heat and mass transfer analogy. The distribution pattern of local Nusselt numbers on single circular tube with heat transfer promoters is similar to that without the heat transfer promoter, but average Nusselt numbers are greatly increased. In case of tube banks without the heat transfer promoter, the Nusselt numbers are much lower in the first row than those of other rows, but the local heat transfer coefficients on all rows are equalized when the heat transfer promoter is installed.

• Journal of Korea Multimedia Society
• /
• v.21 no.10
• /
• pp.1195-1202
• /
• 2018

Artificial neural network is widely used for its excellent performance and implementability. However, traditional neural network needs to learn the system from scratch, with the addition of new input data, the variation of the observation environment, or the change in the form of input/output data. To resolve such a problem, the technique of transfer learning has been proposed. Transfer learning constructs a newly developed target system partially updating existing system and hence provides much more efficient learning process. Until now, transfer learning is mainly studied in the field of image processing and is not yet widely employed in acoustic data processing. In this paper, focusing on the scalability of transfer learning, we apply the concept of transfer learning to the problem of guitar chord classification and evaluate its performance. For this purpose, we build a target system of convolutional neutral network (CNN) based 48 guitar chords classification system by applying the concept of transfer learning to a source system of CNN based 24 guitar chords classification system. We show that the system with transfer learning has performance similar to that of conventional system, but it requires only half the learning time.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.3
• /
• pp.909-914
• /
• 2024

This study suggests a quantitative evaluation of transfer learning, which is widely used in various AI fields, including image recognition for robot vision. Quantitative and qualitative analyses of results applying transfer learning are presented, but transfer learning itself is not discussed. Therefore, this study proposes a quantitative evaluation of transfer learning itself based on MNIST, a handwritten digit database. For the reference network, the change in recognition accuracy according to the depth of the transfer learning frozen layer and the ratio of transfer learning data and pre-training data is tracked. It is observed that when freezing up to the first layer and the ratio of transfer learning data is more than 3%, the recognition accuracy of more than 90% can be stably maintained. The transfer learning quantitative evaluation method of this study can be used to implement transfer learning optimized according to the network structure and type of data in the future, and will expand the scope of the use of robot vision and image analysis AI in various environments.

• Journal of Distribution Science
• /
• v.14 no.11
• /
• pp.49-60
• /
• 2016

Purpose - This paper presents how Multinational Enterprises (MNEs) operate in different tax jurisdiction could decide on its transfer pricing strategy as the optimal solution to increase their global after tax income through transfer pricing and solve their related transfer pricing issues related to distribution cost, consumer, and wholesale vendor. It has been strategy issues for an MNEs to locate its tax basis of wholesale vendor and buyer in a jurisdiction where effective rather low Research design, data, and methodology - The collection of information and data for this research project gathered from various sources of secondary data. The findings of these relevant research topic article and journal were the main source of references for this research project Results - The achievement of management's operational and financial objectives depends on transfer pricing policies availability that is consistent and supports both vendor, wholesaler, distributor and ensuring sufficient documentation and data is available to support the application and arriving at the arm length. Conclusions - The study concluded with an emphasis on the importance of web-designed information about international taxation rules and transfer pricing policy and pricing agreement among wholesale vendor and whole buyer around the world.

• Korean Journal of Air-Conditioning and Refrigeration Engineering
• /
• v.16 no.6
• /
• pp.538-545
• /
• 2004

This paper presents the experimental data for the heat transfer and pressure drop characteristics obtained during the gas cooling process of carbon dioxide in a horizontal tube. The tube in which carbon dioxide flows is made of copper with an inner diameter of 7.73 mm. Experiments were conducted for various mass fluxes and inlet pressures of carbon dioxide. Mass fluxes are controlled at 225, 337 and 450 kg/$m^2$s and inlet pressures are adjust-ed from 7.5 to 8.8 ㎫. The experimental results in this study are compared with the existing correlations for the supercritical heat transfer coefficient, which generally under-predict the measured data. Pressure drop data agree very well with those calculated by the Blasius' equation. Based on the experimental data, a new empirical correlation to estimate the near-critical heat transfer coefficients has been developed.