• Journal of the Korean Society of Radiology
• /
• v.14 no.6
• /
• pp.773-780
• /
• 2020

With the development of the 4th industrial, research is being conducted to prevent diseases and reduce damage in various fields of science and technology such as medicine, health, and bio. As a result, artificial intelligence technology has been introduced and researched for image analysis of radiological examinations. In this paper, we will directly apply a deep learning model for classification and detection of pneumonia using chest X-ray images, and evaluate whether the deep learning model of the Inception series is a useful model for detecting pneumonia. As the experimental material, a chest X-ray image data set provided and shared free of charge by Kaggle was used, and out of the total 3,470 chest X-ray image data, it was classified into 1,870 training data sets, 1,100 validation data sets, and 500 test data sets. I did. As a result of the experiment, the result of metric evaluation of the Inception V3 deep learning model was 94.80% for accuracy, 97.24% for precision, 94.00% for recall, and 95.59 for F1 score. In addition, the accuracy of the final epoch for Inception V3 deep learning modeling was 94.91% for learning modeling and 89.68% for verification modeling for pneumonia detection and classification of chest X-ray images. For the evaluation of the loss function value, the learning modeling was 1.127% and the validation modeling was 4.603%. As a result, it was evaluated that the Inception V3 deep learning model is a very excellent deep learning model in extracting and classifying features of chest image data, and its learning state is also very good. As a result of matrix accuracy evaluation for test modeling, the accuracy of 96% for normal chest X-ray image data and 97% for pneumonia chest X-ray image data was proven. The deep learning model of the Inception series is considered to be a useful deep learning model for classification of chest diseases, and it is expected that it can also play an auxiliary role of human resources, so it is considered that it will be a solution to the problem of insufficient medical personnel. In the future, this study is expected to be presented as basic data for similar studies in the case of similar studies on the diagnosis of pneumonia using deep learning.

• Journal of Korean Society of Disaster and Security
• /
• v.12 no.2
• /
• pp.73-82
• /
• 2019

Recently, as the occurrence frequency of sudden floods due to climate change increased, the flood damage on riverside social infrastructures was extended so that there has been a threat of overflow. Therefore, a rapid prediction of potential flooding in riverside social infrastructure is necessary for administrators. However, most current flood forecasting models including hydraulic model have limitations which are the high accuracy of numerical results but longer simulation time. To alleviate such limitation, data driven models using artificial neural network have been widely used. However, there is a limitation that the existing models can not consider the time-series parameters. In this study the water surface elevation of the Hangang River bridge was predicted using the NARX model considering the time-series parameter. And the results of the ANN and RNN models are compared with the NARX model to determine the suitability of NARX model. Using the 10-year hydrological data from 2009 to 2018, 70% of the hydrological data were used for learning and 15% was used for testing and evaluation respectively. As a result of predicting the water surface elevation after 3 hours from the Hangang River bridge in 2018, the ANN, RNN and NARX models for RMSE were 0.20 m, 0.11 m, and 0.09 m, respectively, and 0.12 m, 0.06 m, and 0.05 m for MAE, and 1.56 m, 0.55 m and 0.10 m for peak errors respectively. By analyzing the error of the prediction results considering the time-series parameters, the NARX model is most suitable for predicting water surface elevation. This is because the NARX model can learn the trend of the time series data and also can derive the accurate prediction value even in the high water surface elevation prediction by using the hyperbolic tangent and Rectified Linear Unit function as an activation function. However, the NARX model has a limit to generate a vanishing gradient as the sequence length becomes longer. In the future, the accuracy of the water surface elevation prediction will be examined by using the LSTM model.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.33 no.5
• /
• pp.455-462
• /
• 2000

The objective of this study was to examine the physicochemical characteristics of coagulation reaction between loess and red tide organisms (RTO) and its feasibility, in developing a technology for the removal of RTO bloom in coastal sea. The physicochemical characteristics of loess were examined for a particle size distribution, surface characteristics by scanning electron microscope, zeta potential, and alkalinity and pH variations in sea water. Two kinds of RTO that were used in this study, Cylindrothen closterium and Skeietonema costatum, were sampled in Masan bay and were cultured in laboratory. Coagulation experiments were conducted using various concentrations of loess, RTO, and a jar tester. The supernatant and RTO culture solution were analyzed for pH, alkalinity, RTO cell number. A negative zeta potential of loess increased with increasing pH at $10^(-3)M$ NaCl solution and had -71.3 mV at pH 9.36. Loess had a positive zeta potential of +1,8 mV at pH 1.98, which resulted in a characteristic of material having an amphoteric surface charge. In NaCl and $CaCl_2$, solutions, loess had a decreasing negative zeta potential with increasing $Na^+\;and\;Ca^(+2)$ ion concentration and then didn't result in a charge reversal due to not occurring specific adsorption for $Na^+$ ion while resulted in a charge reversal due to occurring specific adsorption for $Ca^(+2)$ ion. In sea water, loess and RTO showed the similar zeta potential values of -112,1 and -9.2 mV, respectively and sea sand powder showed the highest zeta potential value of -25.7 mV in the clays. EDLs (electrical double-layers) of loess and RTO were extremely compressed due to high concentration of salts included in sea water, As a result, there didn't almost exist EDL repulsive force between loess and RTO approaching each other and then LVDW (London-yan der Waals) attractive force was always larger than EDL repulsive force to easily form a floe. Removal rates of RTO exponentially increased with increasing a loess concentration. The removal rates steeply increased until $800 mg/l$ of loess, and reached $100{\%}$ at 6,400 mg/l of loess. Removal rates of RTO exponentially increased with increasing a G-value. This indicated that mixing (i.e., collision among particles) was very important for a coagulation reaction. Loess showed the highest RTO removal rates in the clays.

• Korean Journal of Agricultural Science
• /
• v.3 no.2
• /
• pp.145-151
• /
• 1976

This study was conducted to determine the effect of slaked lime and straw used on the soil pH in the flooded condition and yield of rice grown in AKIOCHI paddy field and their residual effects on the rice plants. The results obtained were summarized as follow: when lime and straw were applied, there was on the average 41% of yield increase over plots treated with three elements of chemical fertilizers. When lime plus straw were used, the growth rate at later stage of rice plant was prominent. Damage due to helminthosporium and blast were found less, the rate of lower-leaf death was low, and grain number, per head, filled grain ratio, and weight of rice grain were higher than control. When lime plus straw were used, higher amount of silicate, calcium, nitrogen and potassium was found in the plants at heading stage. The residual effects of lime plus straw were 20% in the first year, about 10% in the second year and 5% in the third year, respectively. Soil pH was affected by both straw and slaked lime, and it was fixed about 8 days after applying in the flooded condition. The following formulae was suggested from the results in the flooded conditions. $$pH=5.5293+8.6007X_1+2.7836X_2-{6.7422X_1}^2-{1.8522X_2}^2-7.000X_1X_2$$ ($X_1$=slaked lime, $X_2$=straw)

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.19 no.1
• /
• pp.66-75
• /
• 2014

This study aims to understand environmental factors that determine spatial distribution of macrozoobenthic community in the southern area (ca 100-500 m depth) of East Sea, Korea, known as a candidate site for carbon storage under the seabed. From sixteen locations sampled in the summer of 2012, a total of 158 species were identified, showing density of $843indiv/m^2$ and biomass of $26.2g\;WW/m^2$, with increasing faunal density towards biologically higher diverse locations. Principal component analysis showed that a total of 33 environmental parameters were reduced to three principal components (PC), indicating sediment, bottom water, and depth, respectively. As sand content was increasing, number of species increased but biomass decreased. Six dominant species including two bivalve species favored high concentrations of ${\Omega}$ aragonite and ${\Omega}$ calcite, indicating that the corresponding species can be severely damaged by ocean acidification or $CO_2$ effluent. Cluaster analysis based on more than 1% density dominant species classified the entire study area into four faunal assemblage (location groups), which were delineated by characteristic species, including (A) Ampelisca miharaensis, (B) Edwardsioides japonica, (C) Maldane cristata, (D) Spiophanes kroeyeri, and clearly separated in terms of geography, bottom water and sediment environment. Overall, a discriminant function model was developed to predict four faunal assemblages from five simply-measured environmental variables (depth, sand content in sediment, temperature, salinity and pH in bottom water) with 100% accuracy, implying that benthic faunal assemablages are closed linked to certain combinations of abiotic factors.