• Journal of Wetlands Research
• /
• v.26 no.2
• /
• pp.147-159
• /
• 2024

Lack of streamflow observations makes model calibration difficult and limits model performance improvement. Satellite-based remote sensing products offer a new alternative as they can be actively utilized to obtain hydrological data. Recently, several studies have shown that artificial intelligence-based solutions are more appropriate than traditional conceptual and physical models. In this study, a data-driven approach combining various recurrent neural networks and decision tree-based algorithms is proposed, and the utilization of satellite remote sensing information for AI training is investigated. The satellite imagery used in this study is from MODIS and SMAP. The proposed approach is validated using publicly available data from 25 watersheds. Inspired by the traditional regionalization approach, a strategy is adopted to learn one data-driven model by integrating data from all basins, and the potential of the proposed approach is evaluated by using a leave-one-out cross-validation regionalization setting to predict streamflow from different basins with one model. The GRU + Light GBM model was found to be a suitable model combination for target basins and showed good streamflow prediction performance in ungauged basins (The average model efficiency coefficient for predicting daily streamflow in 25 ungauged basins is 0.7187) except for the period when streamflow is very small. The influence of satellite remote sensing information was found to be up to 10%, with the additional application of satellite information having a greater impact on streamflow prediction during low or dry seasons than during wet or normal seasons.

• Korean Journal of Child Education & Care
• /
• v.18 no.3
• /
• pp.21-34
• /
• 2018

Objective: The purpose of this study is to investigate the effects of maternal parenting efficacy and emotional expression on infant's social emotional development. Methods: The subjects of this study were 269 infants and their mothers who attended day care centers in G district of S city and the data was collected by survey. The collected data were analyzed by frequency analysis, t-test and ANONA test using SPSS 16.0 program, $Scheff{\acute{e}}$ test by post test, correlation analysis and multiple regression analysis. Results: First, parenting efficacy, emotional expressiveness, and social emotional development of infants showed differences according to age, participation time, and academic background. Parenting efficacy showed a significant difference in age and parenting time, and emotional expressiveness showed significant difference in education and parenting time. Infant's social emotional development showed a significant difference in educational attainment and parenting time. Second, maternal parenting efficacy and emotional expressiveness were found to have an influence on infant's social emotional development. Conclusion/Implications: Maternal parenting efficacy and emotional expressiveness seem to influence the development of social emotions in infants, and these results suggest that, in order to grow the children healthy in social, emotional aspect, various education programs which could improve parenting efficacy and positive emotional expression are necessary.

• Journal of Digital Convergence
• /
• v.13 no.11
• /
• pp.241-254
• /
• 2015

This study examined changes in health behavior and prevalence of hypertension and diabetes during five years and analyzed determinants affecting on geographic variations of them. Data from Korean Community Health Survey in the period of 2008 and 2013 with 246 small districts were analyzed. Data were analyzed using convergence tools such as geographic information system tool and decision tree. During the five years period, areas of the increases in smoking and drinking were southwest regions showed increased smoking and areas of increases in physical activity are western regions. Areas of the increases in the prevalence of hypertension were west and south regions and in the prevalence of diabetes were east and north regions. Determinants affecting on regional variations in the prevalence of hypertension and diabetes were drinking, physical activity, obesity, arthritis, depressive symptom and stress. Mental health program should be developed for non-communicable disease. Thus, to decrease the prevalence of hypertension and diabetes, our study emphasized the necessity to develop customized mental health policies according to the region-specific characteristics.

• Journal of the Korean Institute of Gas
• /
• v.20 no.5
• /
• pp.9-16
• /
• 2016

Buried pipe system is subject to leak or rupture due to internal and external defects with age. Especially, if the pipeline is designed for pressurized gas, the leak can wreak a devastating on its surrounding area. The current method of setting up underground gas pipeline is based on OGP criteria of applying one tenth of the inner pipe pressure. The criteria is applied irrespective of their burial depth or pipe's properties. At times, even the whole safety measures are totally ignored. Considering the magnitude of possible damage from a gas leakage, a precise analytical tool for the risk assessment is urgently needed. The study was conducted to assess possible scenarios of gas accidents and to develop a computer model to minimize the damage. The data from ETA was analyzed intensively, and the model was developed. The model is capable of predicting jet fire influence area with comprehensive input parameters, such as burial depth. The model was calibrated and verified by the historic accident data from Edison Township, New Jersey, the United States. The statistical model was also developed to compare the results of the model in this study and the existing OGP model. They were in good agreement with respect to damage predictions, such as radiation heat coming from 10 meters away from the heat source of gas flame.

• Journal of Korean Society of Forest Science
• /
• v.105 no.3
• /
• pp.330-335
• /
• 2016

This study was conducted to derive merchantable volume ratio for 5 major species such as Pinus Densiflora (Central Region). The data used for this study was from at least more than 1,300 trees of research data throughout the country. the study applied two estimation equations, which were the estimation equation for wood volume ratio representing total wood volume to total tree stem volume and the estimation equation for merchantability representing ratio of merchantable volume to total wood volume. The merchantable volume ratio was derived by multiplying those two estimation equations. In order to gain wood volume ratio(W) from DBH, $W=\frac{a_1}{1+a_2/D}+\frac{b_1}{1+b_2/D}$ model was used. Fitness index of it was more than 99% by species, and other test statistics also indicated the suitability of this equation enough. Merchantability (M) for wood volume applied $M=e^{a_1\(\frac{d}{D}\)^{a_2}}-(b_0+b_1D+b_2D^2+b_3D^3)$ model and fitness index was more than 96% by species. Merchantable volume ratio was assessed using those two estimation equations by each 5 species, and constructed a merchantable volume ratio table. In result, merchuntable volume ratio was little difference between stand types, but there was slightly different with the existing standard such as conifers of 85% and non-conifers of 70%.

• Journal of Korean Society of Forest Science
• /
• v.106 no.4
• /
• pp.450-456
• /
• 2017

This study was conducted to develop estimated equation for mortality rates (volume of dead trees, %) on coniferous and broad-leaved forests, representative forest types of South Korea. There were 6 equation models applied for estimating mortality such as a exponential equation, a Hamilton equation and variables using were DBH, basal area, and site index. Raw data used for estimating mortality were $5^{th}$ and $6^{th}$ national forest inventory data, and mortality was calculated with the difference of stocks between lived trees and dead trees by each sample plots. The most applicable equation to describe mortality on coniferous forest and broad-leaved forest was indicated as $P=(1+e^{(a+b{\times}DBH+c{\times}BA+d{\times}no\_ha+e{\times}density)})^{-1}$ and their goodness of fit showed 34% and 51% respectively. Goodness of fit in both equations were not much high because there were various factors which affect the mortality such as topographic conditions, soil characteristic, climatic factors, site quality, and competition. Therefore, it is considered that explaining mortality in forest with only 2 or 3 variables like DBH, basal area used in this analysis could be very difficult facts. However, this study is certainly worth in that there is no useful information on mortality by each forest type throughout the country at the present, and we would make an effort to promote the fitness of estimated equation for mortality adding competition index, tree crown density etc.

• Journal of Korea Multimedia Society
• /
• v.6 no.1
• /
• pp.148-160
• /
• 2003

TMA(Tissue Mineral Analysis) is very popular method in hair mineral analysis for health care professionals in over 48 countries medical center. Assesment of nutritional minerals and toxic elements in the hair is very important not only for determining adequacy, deficiencies and unbalance, but also for assessing their relative relationships in a body. In Korea, there are some problems in TMA method. Because of not haying a medical information database which is suitable for korean to do analyze, the requested TMA has to send to TEI-USA. However, as the TMA results from TEI-USA is composed of English documents and graphic files prohibited to open, its usability is very low and a lot of dollars has to be payed. Also, it can make some problems in the reliability of the TMA results, since the TMA results are based on the database of western health and mineral standards, To solve these problems, I developed the first Medical Information System of TMA in Korea here. The system can analyze the complex tissue mineral data with multiple stage decision tree classifier. It is also constructed with multiple fuzzy database and hence analyze the TMA data by fuzzy inference methods. The effectiveness test of this systems can be shown the increased business efficiency and satisfaction rate 86% and 92% respectively.

• Journal of Intelligence and Information Systems
• /
• v.20 no.4
• /
• pp.43-58
• /
• 2014

As the use of trading systems has increased rapidly, many researchers have become interested in developing effective stock market prediction models using artificial intelligence techniques. Stock market prediction involves multifaceted interactions between market-controlling factors and unknown random processes. A successful stock prediction model achieves the most accurate result from minimum input data with the least complex model. In this research, we develop a combination model of ${\pi}$-fuzzy logic and support vector machine (SVM) models, using a genetic algorithm to optimize the parameters of the SVM and ${\pi}$-fuzzy functions, as well as feature subset selection to improve the performance of stock market prediction. To evaluate the performance of our proposed model, we compare the performance of our model to other comparative models, including the logistic regression, multiple discriminant analysis, classification and regression tree, artificial neural network, SVM, and fuzzy SVM models, with the same data. The results show that our model outperforms all other comparative models in prediction accuracy as well as return on investment.

• Atmosphere
• /
• v.25 no.2
• /
• pp.221-233
• /
• 2015

Pollen is closely related to health issues such as allergenic rhinitis and asthma as well as intensifying atopic syndrome. Information on current and future spatio-temporal distribution of allergenic pollen is needed to address such issues. In this study, the Community Multiscale Air Quality Modeling (CMAQ) was utilized as a base modeling system to forecast pollen dispersal from oak trees. Pollen emission is one of the most important parts in the dispersal modeling system. Areal emission factor was determined from gridded areal fraction of oak trees, which was produced by the analysis of the tree type maps (1:5000) obtained from the Korea Forest Service. Daily total pollen production was estimated by a robust multiple regression model of weather conditions and pollen concentration. Hourly emission factor was determined from wind speed and friction velocity. Hourly pollen emission was then calculated by multiplying areal emission factor, daily total pollen production, and hourly emission factor. Forecast data from the KMA UM LDAPS (Korea Meteorological Administration Unified Model Local Data Assimilation and Prediction System) was utilized as input. For the verification of the model, daily observed pollen concentration from 12 sites in Korea during the pollen season of 2014. Although the model showed a tendency of over-estimation in terms of the seasonal and daily mean concentrations, overall concentration was similar to the observation. Comparison at the hourly output showed distinctive delay of the peak hours by the model at the 'Pocheon' site. It was speculated that the constant release of hourly number of pollen in the modeling framework caused the delay.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.9 no.3
• /
• pp.51-58
• /
• 2006

The purpose of this research presents a method to position and makes the structure for eco-corridors reasonably with collectable analysing results of various effects shown in mammals' road-kill at 429 points. Target animals of this research are Leopard cat, Siberian weasel, Raccoon dog, Korean hare, Eurasian red squirrel, Siberian chipmunk and Water deer. The results derived from the empirical analysis on the contents above are followed. First, according to the results as for Leopard cat road kill analysis, which is designated as Endangered Species Class II, the eco-corridor might be located at near village having stead food in order to decrease the frequencies of road-kill, because its road kill points were mainly collected at 4 lane hilly road with mountain-road-farm area geological type of. Second, because Siberian weasel's road kill was detected at 2 lane hilly road with mountain-road-stream geological type, the eco-corridor might be located at near a mill to decrease road-kill frequencies. Third, the road-kill frequency of Eurasian red squirrel can be reduced when the eco-corridor is located at the area across coniferous tree near 4 lane west sea freeway with mountain-road-mountain. Fourth, the road-kill of Raccoon dog can be reduced when the eco-corridor is located at 4 lane mountain road or hilly road with the geological type having farm land-road-mountain(stream). Fifth, Korean hare's road-kill can be reduced when the eco-corridor is located at grass land across ridge line of mountain, because wild rabbit road kill was happened at 4 lane mountain road or 2 lane mountain road(mountain-road-mountain). Sixth, As for Siberian chipmunk, the eco-corridor might be located at the side slope of mountain road at 2 lane mountain road under the speed of 60km/h with mountain-road-mountain. Seventh, For Water deer, the eco-corridor might be located at 4 lane hilly road with mountain-road-farm land. As for Common otter, Amur hedgehog, Yellow-throated marten, Weasel, it is difficult to specify the proper site of eco-corridor due to the lack of data. Eco-corridors for carnivores might be well located at 4 lane hilly road or 2 lane hilly road with mountain-road-farm land, and the track for herbivores might be well located as a overhead bridge on mountain-road-mountain type across mountains. In order to position eco-corridors for wildlife properly, we have to research animal's behavior with ecological background, and to consider the local uniqueness and regularly collect the empirical road-kill data in long term 3 to 5 year, which can be the foundation for the more suitable place of wild life eco-corridors.