• Journal of Digital Convergence
• /
• v.18 no.2
• /
• pp.261-268
• /
• 2020

Machine learning is achieved by making a cost function from a given digital signal data and optimizing the cost function. The cost function here has local minimums in the cost function depending on the amount of digital signal data and the structure of the neural network. These local minimums make a problem that prevents learning. Among the many ways of solving these methods, our proposed method is to change the learning step-size. Unlike existed methods using the learning rate (step-size) as a fixed constant, the use of multivariate function as the cost function prevent unnecessary machine learning and find the best way to the minimum value. Numerical experiments show that the results of the proposed method improve about 3%(88.8%→91.5%) performance using the proposed method rather than the existed methods.

• Weed & Turfgrass Science
• /
• v.3 no.1
• /
• pp.50-55
• /
• 2014

This study is aimed at classifying the syntaxa of Ceratopteris thalictroides dominant community on the Nakdong River, and to collect basic data for research of habitat. The communities were carried out by using the Z.-M. School's method and numerical classification technique. The result of syntaxa was classified three communities such as Persicaria japonica-Ceratopteris thalictroides community, Lindernia procumbens-Ceratropteris thalictroides community, and Limnophila indica-Ceratopteris thalictroides community. The ordination analysis displayed the vegetation types with respect to complex environmental gradients. After ordination and clustering analysis, the effective humidity, soil stability, trampling effects, anthropogenic effects and flooding frequency were identified as the important factors deciding the vegetation pattern. It was pointed out to establish a long-term ecological site for protecting such vulnerable vegetation against overexploitation and global climate change.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.13 no.1
• /
• pp.37-49
• /
• 2000

This paper is for the first essential study on the development of unified finite element formulations for solving problems related to the dynamics/thermoelastics behavior of solids. In the first part of formulations, the finite element method is based on the introduction of a new quantity defined as heat displacement, which allows the heat conduction equations to be written in a form equivalent to the equation of motion, and the equations of coupled thermoelasticity to be written in a unified form. The equations obtained are used to express a variational formulation which, together with the concept of generalized coordinates, yields a set of differential equations with the time as an independent variable. Using the Laplace transform, the resulting finite element equations are described in the transform domain. In the second, the Laplace transform is applied to both the equation of heat conduction derived in the first part and the equations of motions and their corresponding boundary conditions, which is referred to the transformed equation. Selections of interpolation functions dependent on only the space variable and an application of the weighted residual method to the coupled equation result in the necessary finite element matrices in the transformed domain. Finally, to prove the validity of two approaches, a comparison with one finite element equation and the other is made term by term.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2008.04a
• /
• pp.697-700
• /
• 2008

The object of this paper is to propose a logical prediction model of a concrete creep using rheology. Rheology is the study on the flow and stress relationship of matter under the influence of an applied stress. It is also estimated as an effective theory to describe concrete long-term deformations. According to a time dependency and a mechanism of occurrence, the proposed creep model was divided into four components, such as an elastic deformation, a long-term creep, a time dependent short-term creep and a time independent short-term creep. Evaluation on an actual creep deformation pattern by time passage confirmed these classification. In order to approve a rationality of the proposed model, most coefficients of each components were derived by the microprestresssolidification theory and design codes. Numerical approaches were also used when it was restricted within narrow limits. Finally, the proposed rheolgical model was verified by actual creep test results and compared with common methods.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2011.06a
• /
• pp.34-36
• /
• 2011

ENCs is edited based on the numerical charts for publishing paper charts and serviced in forms of grid styles. For this reason, the density of sounding information of ENCs is not consistent and was required for improvement. In this study, K-Means, ISODATA clustering algorithm as classification methods for satellite image was reviewed and adopted to case study. The developed results include loading module of ENC data, improvement algorithm of sounding information, writing module of ENC data. According to the results of algorithm, we could confirm the improved result.

• Journal of Digital Convergence
• /
• v.15 no.7
• /
• pp.213-222
• /
• 2017

This paper proposes a new service design which is deep learning-based image retrieval system for product search on O2O shopping mall platform. We have implemented deep learning technology that provides more convenient retrieval service for diverse images of many products that are sold in the internet shopping malls. In order to implement this retrieval system, real data used by shopping mall companies were used as experimental data. However, result from several experiments have confirmed deterioration of retrieval performance due to data components. In order to improve the performance, the learning data that interferes with the retrieval is revised several times, and then the values of experimental result are quantified with the verification data. Using the numerical values of these experiments, we have applied them to the new service design in this system.

• Journal of the Korean GEO-environmental Society
• /
• v.17 no.1
• /
• pp.21-28
• /
• 2016

This paper deals with contents on the estimation of rail beam from the geotechnical engineering aspect. Rail beam is reinforced rail installed on the inside and outside of rail to prevent differential settlement during the construction period of railroad crossing construction. Such rail beam is frequently being installed to ensure stability of existing railroad facilities because of increasing constructions of underground structures crossing railroad in recent. However, there is a difficulty in design due to lack of design standard on rail beam length. Furthermore, derailing accidents are also occurring as a result of rail beam length shortage. Accordingly, this paper presented flow chart based on the classification into soil ground and bedrock ground for the rail beam length estimation. In addition, case study was conducted on rail combination and location through which effective rail combination and location were ensured.

• Journal of the Korean Society of Safety
• /
• v.26 no.3
• /
• pp.91-97
• /
• 2011

The weight of elderly people in Korea has been increasing. Statistics show that the percentage of the elderly people in Korea was 3.1% in 1970; 3.8% in 1980; 5.1% in 1990, and 7.2% in 2000. Based on this trend, thus, the number of elderly people could be estimated to be 14% of the whole Korean population in 2018. This reveals that Korea is entering a super-aging society with remarkable fast pace. In such a change, the statistics related to elderly people driving license and the occurrence of traffic accidents are showing a noticeable numerical value. The number of traffic accident fatality in Korea ranks the highest value in OECD Countries. However, the research on old drivers in the nation is going on partially centering on system improvement and management scheme. Thus, first of all, researches about the linkage & characteristics between the driving behavior of old drivers and traffic accident should be implemented, in order properly to draw system improvement and management scheme for the old drivers. Therefore, the focus of this study is the influence model for discerning the severity of the age-old-caused traffic accidents by inquiring into the relation between the Driving Aptitude Test items that make it possible to measure their behavioral characteristics and influential factors by age group on the basis of the data on traffic accidents. The analysis results can be used as basic data for suggesting the behavioral research and countermeasure for traffic safety and its management for old driver in preparation for the aging society.

• Journal of Digital Convergence
• /
• v.19 no.10
• /
• pp.253-263
• /
• 2021

Machine learning is the process of constructing a cost function using learning data used for learning and an artificial neural network to predict the data, and finding parameters that minimize the cost function. Parameters are changed by using the gradient-based method of the cost function. The more complex the digital signal and the more complex the problem to be learned, the more complex and deeper the structure of the artificial neural network. Such a complex and deep neural network structure can cause over-fitting problems. In order to avoid over-fitting, a weight decay regularization method of parameters is used. We additionally use the value of the cost function in this method. In this way, the accuracy of machine learning is improved, and the superiority is confirmed through numerical experiments. These results derive accurate values for a wide range of artificial intelligence data through machine learning.

• The Journal of the Korea Contents Association
• /
• v.19 no.3
• /
• pp.261-274
• /
• 2019

This paper aims to develop an appropriate methodology for establishing an investment strategy for 'demonstration of artificial rainfall technology using UAV' and that include establishment of a technology classification, set of indicators for technology evaluation, suggestion of final key technology as a whole study area. It is designed to complement the latest research trend analysis results and expert committee opinions using quantitative analysis. The key indicators for technology evaluation consisted of three major items (activity, technology, marketability) and 10 detailed indicators. The AHP questionnaire was conducted to analyze the importance of indicators. As a result, it was analyzed that the attribute of the technology itself is most important, and the order of closeness to the implementation of the core function (centrality), feasibility (feasibility). Among the 16 technology groups, top investment priority groups were analyzed as ground seeding, artificial rainfall verification, spreading and diffusion of seeding material, artificial rainfall numerical modeling, and UAV sensor technology.