• The Journal of Engineering Geology
• /
• v.29 no.1
• /
• pp.51-68
• /
• 2019

This study performs to evaluate adaptability of seashore filtering type seawater-intake which adapts dua1 filter well alternative for direct seawater-intake. This study varies filter condition of seashore free surface aquifer which is composed of sand layer then installs real size dual filter well and single filter well to evaluate water permeability and proper pumping amount according to filter condition. According to result of step aquifer test, it is analysed that 110.3% synergy effect of water permeability coefficient is happened compare to single filter since dual filter well has better improvement. dual filter has higher water permeability coefficient compare to same pumping amount, this means dual filter has more improved water permeability than single filter. According to analysis result of continuous aquifer test, it is evaluated that dual filter well (SD1200) has higher water permeability than single filter well (SS800) by analysis of water permeability coefficient using monitoring well and gauging well, it is also analysed dual filter has 110.7% synergy effect of water permeability coefficient. As a evaluation result of pumping amount according to analysis of water level dropping rate, it is analysed that dual filter well increased 122.8% pumping amount compare to single filter well when water level dropping is 2.0 m. As a result of calculating proper pumping amount using water level dropping rate, it is analysed that dual filter well shows 136.0% higher pumping amount compare to single filter well. It is evaluated that proper pumping amount has 122.8~160% improvement compare to single filter, pumping amount improvement rate is 139.6% compare to averaged single filter. In other words, about 40% water intake efficiency can be improved by just installation of dual filter compare to normal well. Proper pumping amount of dual filter well using inflection point is 2843.3 L/min and it is evaluated that daily seawater intake amount is about $4,100m^3/day$ (${\fallingdotseq}4094.3m^3/day$) in one hole of dual filter well. Since it is possible to intake plenty of water in one hole, higher adaptability is anticipated. In case of intaking seawater using dual filter well, no worries regarding damages on facilities caused by natural disaster such as severe weather or typhoon, improvement of pollution is anticipated due to seashore sand layer acts like filter. Therefore, It can be alternative of environmental issue for existing seawater intake technique, can save maintenance expenses related to installation fee or damages and has excellent adaptability in economic aspect. The result of this study will be utilized as a basic data of site demonstration test for adaptation of riverside filtered water of upcoming dual filter well and this study is also anticipated to present standard of well design and construction related to riverside filter and seashore filter technique.

• Journal of the Korean Geotechnical Society
• /
• v.35 no.5
• /
• pp.5-19
• /
• 2019

Considering the natural phenomenon in which steep slopes ($65^{\circ}{\sim}85^{\circ}$) consisting of rock mass remain stable for decades, slopes steeper than 1:0.5 (the standard of slope angle for blast rock) may be applied in geotechnical conditions which are similar to those above at the design and initial construction stages. In the process of analysing the stability of a good to fair continuum rock slope that can be designed as a steep slope, a general method of estimating rock mass strength properties from design practice perspective was required. Practical and genealized engineering methods of determining the properties of a rock mass are important for a good continuum rock slope that can be designed as a steep slope. The Genealized Hoek-Brown (H-B) failure criterion and GSI (Geological Strength Index), which were revised and supplemented by Hoek et al. (2002), were assessed as rock mass characterization systems fully taking into account the effects of discontinuities, and were widely utilized as a method for calculating equivalent Mohr-Coulomb shear strength (balancing the areas) according to stress changes. The concept of calculating equivalent M-C shear strength according to the change of confining stress range was proposed, and on a slope, the equivalent shear strength changes sensitively with changes in the maximum confining stress (${{\sigma}^{\prime}}_{3max}$ or normal stress), making it difficult to use it in practical design. In this study, the method of estimating the strength properties (an iso-angle division method) that can be applied universally within the maximum confining stress range for a good to fair continuum rock mass slope is proposed by applying the H-B failure criterion. In order to assess the validity and applicability of the proposed method of estimating the shear strength (A), the rock slope, which is a study object, was selected as the type of rock (igneous, metamorphic, sedimentary) on the steep slope near the existing working design site. It is compared and analyzed with the equivalent M-C shear strength (balancing the areas) proposed by Hoek. The equivalent M-C shear strength of the balancing the areas method and iso-angle division method was estimated using the RocLab program (geotechnical properties calculation software based on the H-B failure criterion (2002)) by using the basic data of the laboratory rock triaxial compression test at the existing working design site and the face mapping of discontinuities on the rock slope of study area. The calculated equivalent M-C shear strength of the balancing the areas method was interlinked to show very large or small cohesion and internal friction angles (generally, greater than $45^{\circ}$). The equivalent M-C shear strength of the iso-angle division is in-between the equivalent M-C shear properties of the balancing the areas, and the internal friction angles show a range of $30^{\circ}$ to $42^{\circ}$. We compared and analyzed the shear strength (A) of the iso-angle division method at the study area with the shear strength (B) of the existing working design site with similar or the same grade RMR each other. The application of the proposed iso-angle division method was indirectly evaluated through the results of the stability analysis (limit equilibrium analysis and finite element analysis) applied with these the strength properties. The difference between A and B of the shear strength is about 10%. LEM results (in wet condition) showed that Fs (A) = 14.08~58.22 (average 32.9) and Fs (B) = 18.39~60.04 (average 32.2), which were similar in accordance with the same rock types. As a result of FEM, displacement (A) = 0.13~0.65 mm (average 0.27 mm) and displacement (B) = 0.14~1.07 mm (average 0.37 mm). Using the GSI and Hoek-Brown failure criterion, the significant result could be identified in the application evaluation. Therefore, the strength properties of rock mass estimated by the iso-angle division method could be applied with practical shear strength.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.57-73
• /
• 2021

Maintenance and prevention of failure through anomaly detection of ICT infrastructure is becoming important. System monitoring data is multidimensional time series data. When we deal with multidimensional time series data, we have difficulty in considering both characteristics of multidimensional data and characteristics of time series data. When dealing with multidimensional data, correlation between variables should be considered. Existing methods such as probability and linear base, distance base, etc. are degraded due to limitations called the curse of dimensions. In addition, time series data is preprocessed by applying sliding window technique and time series decomposition for self-correlation analysis. These techniques are the cause of increasing the dimension of data, so it is necessary to supplement them. The anomaly detection field is an old research field, and statistical methods and regression analysis were used in the early days. Currently, there are active studies to apply machine learning and artificial neural network technology to this field. Statistically based methods are difficult to apply when data is non-homogeneous, and do not detect local outliers well. The regression analysis method compares the predictive value and the actual value after learning the regression formula based on the parametric statistics and it detects abnormality. Anomaly detection using regression analysis has the disadvantage that the performance is lowered when the model is not solid and the noise or outliers of the data are included. There is a restriction that learning data with noise or outliers should be used. The autoencoder using artificial neural networks is learned to output as similar as possible to input data. It has many advantages compared to existing probability and linear model, cluster analysis, and map learning. It can be applied to data that does not satisfy probability distribution or linear assumption. In addition, it is possible to learn non-mapping without label data for teaching. However, there is a limitation of local outlier identification of multidimensional data in anomaly detection, and there is a problem that the dimension of data is greatly increased due to the characteristics of time series data. In this study, we propose a CMAE (Conditional Multimodal Autoencoder) that enhances the performance of anomaly detection by considering local outliers and time series characteristics. First, we applied Multimodal Autoencoder (MAE) to improve the limitations of local outlier identification of multidimensional data. Multimodals are commonly used to learn different types of inputs, such as voice and image. The different modal shares the bottleneck effect of Autoencoder and it learns correlation. In addition, CAE (Conditional Autoencoder) was used to learn the characteristics of time series data effectively without increasing the dimension of data. In general, conditional input mainly uses category variables, but in this study, time was used as a condition to learn periodicity. The CMAE model proposed in this paper was verified by comparing with the Unimodal Autoencoder (UAE) and Multi-modal Autoencoder (MAE). The restoration performance of Autoencoder for 41 variables was confirmed in the proposed model and the comparison model. The restoration performance is different by variables, and the restoration is normally well operated because the loss value is small for Memory, Disk, and Network modals in all three Autoencoder models. The process modal did not show a significant difference in all three models, and the CPU modal showed excellent performance in CMAE. ROC curve was prepared for the evaluation of anomaly detection performance in the proposed model and the comparison model, and AUC, accuracy, precision, recall, and F1-score were compared. In all indicators, the performance was shown in the order of CMAE, MAE, and AE. Especially, the reproduction rate was 0.9828 for CMAE, which can be confirmed to detect almost most of the abnormalities. The accuracy of the model was also improved and 87.12%, and the F1-score was 0.8883, which is considered to be suitable for anomaly detection. In practical aspect, the proposed model has an additional advantage in addition to performance improvement. The use of techniques such as time series decomposition and sliding windows has the disadvantage of managing unnecessary procedures; and their dimensional increase can cause a decrease in the computational speed in inference.The proposed model has characteristics that are easy to apply to practical tasks such as inference speed and model management.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.139-156
• /
• 2021

The biggest reason for using a deep learning model in image classification is that it is possible to consider the relationship between each region by extracting each region's features from the overall information of the image. However, the CNN model may not be suitable for emotional image data without the image's regional features. To solve the difficulty of classifying emotion images, many researchers each year propose a CNN-based architecture suitable for emotion images. Studies on the relationship between color and human emotion were also conducted, and results were derived that different emotions are induced according to color. In studies using deep learning, there have been studies that apply color information to image subtraction classification. The case where the image's color information is additionally used than the case where the classification model is trained with only the image improves the accuracy of classifying image emotions. This study proposes two ways to increase the accuracy by incorporating the result value after the model classifies an image's emotion. Both methods improve accuracy by modifying the result value based on statistics using the color of the picture. When performing the test by finding the two-color combinations most distributed for all training data, the two-color combinations most distributed for each test data image were found. The result values were corrected according to the color combination distribution. This method weights the result value obtained after the model classifies an image's emotion by creating an expression based on the log function and the exponential function. Emotion6, classified into six emotions, and Artphoto classified into eight categories were used for the image data. Densenet169, Mnasnet, Resnet101, Resnet152, and Vgg19 architectures were used for the CNN model, and the performance evaluation was compared before and after applying the two-stage learning to the CNN model. Inspired by color psychology, which deals with the relationship between colors and emotions, when creating a model that classifies an image's sentiment, we studied how to improve accuracy by modifying the result values based on color. Sixteen colors were used: red, orange, yellow, green, blue, indigo, purple, turquoise, pink, magenta, brown, gray, silver, gold, white, and black. It has meaning. Using Scikit-learn's Clustering, the seven colors that are primarily distributed in the image are checked. Then, the RGB coordinate values of the colors from the image are compared with the RGB coordinate values of the 16 colors presented in the above data. That is, it was converted to the closest color. Suppose three or more color combinations are selected. In that case, too many color combinations occur, resulting in a problem in which the distribution is scattered, so a situation fewer influences the result value. Therefore, to solve this problem, two-color combinations were found and weighted to the model. Before training, the most distributed color combinations were found for all training data images. The distribution of color combinations for each class was stored in a Python dictionary format to be used during testing. During the test, the two-color combinations that are most distributed for each test data image are found. After that, we checked how the color combinations were distributed in the training data and corrected the result. We devised several equations to weight the result value from the model based on the extracted color as described above. The data set was randomly divided by 80:20, and the model was verified using 20% of the data as a test set. After splitting the remaining 80% of the data into five divisions to perform 5-fold cross-validation, the model was trained five times using different verification datasets. Finally, the performance was checked using the test dataset that was previously separated. Adam was used as the activation function, and the learning rate was set to 0.01. The training was performed as much as 20 epochs, and if the validation loss value did not decrease during five epochs of learning, the experiment was stopped. Early tapping was set to load the model with the best validation loss value. The classification accuracy was better when the extracted information using color properties was used together than the case using only the CNN architecture.

• The Journal of Korean Society for Radiation Therapy
• /
• v.30 no.1_2
• /
• pp.9-16
• /
• 2018

Purpose : Brain Stereotactic Radiosurgery can treat non-invasive diseases with high rates of complications due to surgical operations. However, brain stereotactic radiosurgery may be accompanied by radiation induced side effects such as fractionation radiation therapy because it uses radiation. The effects of Coplanar Volumetric Modulated Arc Therapy(C-VMAT) and Non-Coplanar Volumetric Modulated Arc Therapy(NC-VMAT) on surrounding normal tissues were analyzed in order to reduce the side effects caused fractionation radiation therapy such as head and neck. But, brain stereotactic radiosurgery these contents were not analyzed. In this study, we evaluated the usefulness of NC-VMAT by comparing and analyzing C-VMAT and NC-VMAT in patients who underwent brain stereotactic radiosurgery. Methods and materials : With C-VMAT and NC-VMAT, 13 treatment plans for brain stereotactic radiosurgery were established. The Planning Target Volume ranged from a minimum of 0.78 cc to a maximum of 12.26 cc, Prescription doses were prescribed between 15 and 24 Gy. Treatment machine was TrueBeam STx (Varian Medical Systems, USA). The energy used in the treatment plan was 6 MV Flattening Filter Free (6FFF) X-ray. The C-VMAT treatment plan used a half 2 arc or full 2 arc treatment plan, and the NC-VMAT treatment plan used 3 to 7 Arc 40 to 190 degrees. The angle of the couch was planned to be 3-7 angles. Results : The mean value of the maximum dose was $105.1{\pm}1.37%$ in C-VMAT and $105.8{\pm}1.71%$ in NC-VMAT. Conformity index of C-VMAT was $1.08{\pm}0.08$ and homogeneity index was $1.03{\pm}0.01$. Conformity index of NC-VMAT was $1.17{\pm}0.1$ and homogeneity index was $1.04{\pm}0.01$. $V_2$, $V_8$, $V_{12}$, $V_{18}$, $V_{24}$ of the brain were $176{\pm}149.36cc$, $31.50{\pm}25.03cc$, $16.53{\pm}12.63cc$, $8.60{\pm}6.87cc$ and $4.03{\pm}3.43cc$ in the C-VMAT and $135.55{\pm}115.93cc$, $24.34{\pm}17.68cc$, $14.74{\pm}10.97cc$, $8.55{\pm}6.79cc$, $4.23{\pm}3.48cc$. Conclusions : The maximum dose, conformity index, and homogeneity index showed no significant difference between C-VMAT and NC-VMAT. $V_2$ to $V_{18}$ of the brain showed a difference of at least 0.5 % to 48 %. $V_{19}$ to $V_{24}$ of the brain showed a difference of at least 0.4 % to 4.8 %. When we compare the mean value of $V_{12}$ that Radione-crosis begins to generate, NC-VMAT has about 12.2 % less amount than C-VMAT. These results suggest that if NC-VMAT is used, the volume of $V_2$ to $V_{18}$ can be reduced, which can reduce Radionecrosis.

• The Journal of Korean Society for Radiation Therapy
• /
• v.32
• /
• pp.41-52
• /
• 2020

Purpose: To find out the dosimetric usefulness, setup reproducibility and efficiency of applying 3D Bolus by comparing two treatment plans in which Commercial Bolus and 3D Bolus produced by 3D Printing Technology were applied to the neck during VMAT treatment of Hypopahrynx Cancer to evaluate the clinical applicability. Materials and Methods: Based on the CT image of the RANDO phantom to which CB was applied, 3D Bolus were fabricated in the same form. 3D Bolus was printed with a polyurethane acrylate resin with a density of 1.2g/㎤ through the SLA technique using OMG SLA 660 Printer and MaterializeMagics software. Based on two CT images using CB and 3D Bolus, a treatment plan was established assuming VMAT treatment of Hypopharynx Cancer. CBCT images were obtained for each of the two established treatment plans 18 times, and the treatment efficiency was evaluated by measuring the setup time each time. Based on the obtained CBCT image, the adaptive plan was performed through Pinnacle, a computerized treatment planning system, to evaluate target, normal organ dose evaluation, and changes in bolus volume. Results: The setup time for each treatment plan was reduced by an average of 28 sec in the 3D Bolus treatment plan compared to the CB treatment plan. The Bolus Volume change during the pretreatment period was 86.1±2.70㎤ in 83.9㎤ of CB Initial Plan and 99.8±0.46㎤ in 92.2㎤ of 3D Bolus Initial Plan. The change in CTV Min Value was 167.4±19.38cGy in CB Initial Plan 191.6cGy and 149.5±18.27cGy in 3D Bolus Initial Plan 167.3cGy. The change in CTV Mean Value was 228.3±0.38cGy in CB Initial Plan 227.1cGy and 227.7±0.30cGy in 3D Bolus Initial Plan 225.9cGy. The change in PTV Min Value was 74.9±19.47cGy in CB Initial Plan 128.5cGy and 83.2±12.92cGy in 3D Bolus Initial Plan 139.9cGy. The change in PTV Mean Value was 226.2±0.83cGy in CB Initial Plan 225.4cGy and 225.8±0.33cGy in 3D Bolus Initial Plan 224.1cGy. The maximum value for the normal organ spinal cord was the same as 135.6cGy on average each time. Conclusion: From the experimental results of this paper, it was found that the application of 3D Bolus to the irregular body surface is more dosimetrically useful than the application of Commercial Bolus, and the setup reproducibility and efficiency are excellent. If further case studies along with research on the diversity of 3D printing materials are conducted in the future, the application of 3D Bolus in the field of radiation therapy is expected to proceed more actively.

• Journal of Intelligence and Information Systems
• /
• v.27 no.1
• /
• pp.23-46
• /
• 2021

Collaborative filtering, which is often used in personalization recommendations, is recognized as a very useful technique to find similar customers and recommend products to them based on their purchase history. However, the traditional collaborative filtering technique has raised the question of having difficulty calculating the similarity for new customers or products due to the method of calculating similaritiesbased on direct connections and common features among customers. For this reason, a hybrid technique was designed to use content-based filtering techniques together. On the one hand, efforts have been made to solve these problems by applying the structural characteristics of social networks. This applies a method of indirectly calculating similarities through their similar customers placed between them. This means creating a customer's network based on purchasing data and calculating the similarity between the two based on the features of the network that indirectly connects the two customers within this network. Such similarity can be used as a measure to predict whether the target customer accepts recommendations. The centrality metrics of networks can be utilized for the calculation of these similarities. Different centrality metrics have important implications in that they may have different effects on recommended performance. In this study, furthermore, the effect of these centrality metrics on the performance of recommendation may vary depending on recommender algorithms. In addition, recommendation techniques using network analysis can be expected to contribute to increasing recommendation performance even if they apply not only to new customers or products but also to entire customers or products. By considering a customer's purchase of an item as a link generated between the customer and the item on the network, the prediction of user acceptance of recommendation is solved as a prediction of whether a new link will be created between them. As the classification models fit the purpose of solving the binary problem of whether the link is engaged or not, decision tree, k-nearest neighbors (KNN), logistic regression, artificial neural network, and support vector machine (SVM) are selected in the research. The data for performance evaluation used order data collected from an online shopping mall over four years and two months. Among them, the previous three years and eight months constitute social networks composed of and the experiment was conducted by organizing the data collected into the social network. The next four months' records were used to train and evaluate recommender models. Experiments with the centrality metrics applied to each model show that the recommendation acceptance rates of the centrality metrics are different for each algorithm at a meaningful level. In this work, we analyzed only four commonly used centrality metrics: degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality. Eigenvector centrality records the lowest performance in all models except support vector machines. Closeness centrality and betweenness centrality show similar performance across all models. Degree centrality ranking moderate across overall models while betweenness centrality always ranking higher than degree centrality. Finally, closeness centrality is characterized by distinct differences in performance according to the model. It ranks first in logistic regression, artificial neural network, and decision tree withnumerically high performance. However, it only records very low rankings in support vector machine and K-neighborhood with low-performance levels. As the experiment results reveal, in a classification model, network centrality metrics over a subnetwork that connects the two nodes can effectively predict the connectivity between two nodes in a social network. Furthermore, each metric has a different performance depending on the classification model type. This result implies that choosing appropriate metrics for each algorithm can lead to achieving higher recommendation performance. In general, betweenness centrality can guarantee a high level of performance in any model. It would be possible to consider the introduction of proximity centrality to obtain higher performance for certain models.

• The Journal of Korean Society for Radiation Therapy
• /
• v.32
• /
• pp.7-15
• /
• 2020

Purpose: In modern radiotherapy technology, several methods of image guided radiation therapy (IGRT) are used to deliver accurate doses to tumor target locations and normal organs, including CBCT (Cone Beam Computed Tomography) and other devices, ExacTrac System, other than CBCT equipped with linear accelerators. In previous studies comparing the two systems, positional errors were analysed rearwards using Offline-view or evaluated only with a Yaw rotation with the X, Y, and Z axes. In this study, when using CBCT and ExacTrac to perform 6 Degree of the Freedom(DoF) Online IGRT in a treatment center with two equipment, the difference between the set-up calibration values seen in each system, the time taken for patient set-up, and the radiation usefulness of the imaging device is evaluated. Materials and Methods: In order to evaluate the difference between mobile calibrations and exposure radiation dose, the glass dosimetry and Rando Phantom were used for 11 cancer patients with head circumference from March to October 2017 in order to assess the difference between mobile calibrations and the time taken from Set-up to shortly before IGRT. CBCT and ExacTrac System were used for IGRT of all patients. An average of 10 CBCT and ExacTrac images were obtained per patient during the total treatment period, and the difference in 6D Online Automation values between the two systems was calculated within the ROI setting. In this case, the area of interest designation in the image obtained from CBCT was fixed to the same anatomical structure as the image obtained through ExacTrac. The difference in positional values for the six axes (SI, AP, LR; Rotation group: Pitch, Roll, Rtn) between the two systems, the total time taken from patient set-up to just before IGRT, and exposure dose were measured and compared respectively with the RandoPhantom. Results: the set-up error in the phantom and patient was less than 1mm in the translation group and less than 1.5° in the rotation group, and the RMS values of all axes except the Rtn value were less than 1mm and 1°. The time taken to correct the set-up error in each system was an average of 256±47.6sec for IGRT using CBCT and 84±3.5sec for ExacTrac, respectively. Radiation exposure dose by IGRT per treatment was measured at 37 times higher than ExacTrac in CBCT and ExacTrac at 2.468mGy and 0.066mGy at Oral Mucosa among the 7 measurement locations in the head and neck area. Conclusion: Through 6D online automatic positioning between the CBCT and ExacTrac systems, the set-up error was found to be less than 1mm, 1.02°, including the patient's movement (random error), as well as the systematic error of the two systems. This error range is considered to be reasonable when considering that the PTV Margin is 3mm during the head and neck IMRT treatment in the present study. However, considering the changes in target and risk organs due to changes in patient weight during the treatment period, it is considered to be appropriately used in combination with CBCT.

• Journal of Intelligence and Information Systems
• /
• v.27 no.1
• /
• pp.191-207
• /
• 2021

Up to this day, mobile communications have evolved rapidly over the decades, mainly focusing on speed-up to meet the growing data demands of 2G to 5G. And with the start of the 5G era, efforts are being made to provide such various services to customers, as IoT, V2X, robots, artificial intelligence, augmented virtual reality, and smart cities, which are expected to change the environment of our lives and industries as a whole. In a bid to provide those services, on top of high speed data, reduced latency and reliability are critical for real-time services. Thus, 5G has paved the way for service delivery through maximum speed of 20Gbps, a delay of 1ms, and a connecting device of 106/㎢ In particular, in intelligent traffic control systems and services using various vehicle-based Vehicle to X (V2X), such as traffic control, in addition to high-speed data speed, reduction of delay and reliability for real-time services are very important. 5G communication uses high frequencies of 3.5Ghz and 28Ghz. These high-frequency waves can go with high-speed thanks to their straightness while their short wavelength and small diffraction angle limit their reach to distance and prevent them from penetrating walls, causing restrictions on their use indoors. Therefore, under existing networks it's difficult to overcome these constraints. The underlying centralized SDN also has a limited capability in offering delay-sensitive services because communication with many nodes creates overload in its processing. Basically, SDN, which means a structure that separates signals from the control plane from packets in the data plane, requires control of the delay-related tree structure available in the event of an emergency during autonomous driving. In these scenarios, the network architecture that handles in-vehicle information is a major variable of delay. Since SDNs in general centralized structures are difficult to meet the desired delay level, studies on the optimal size of SDNs for information processing should be conducted. Thus, SDNs need to be separated on a certain scale and construct a new type of network, which can efficiently respond to dynamically changing traffic and provide high-quality, flexible services. Moreover, the structure of these networks is closely related to ultra-low latency, high confidence, and hyper-connectivity and should be based on a new form of split SDN rather than an existing centralized SDN structure, even in the case of the worst condition. And in these SDN structural networks, where automobiles pass through small 5G cells very quickly, the information change cycle, round trip delay (RTD), and the data processing time of SDN are highly correlated with the delay. Of these, RDT is not a significant factor because it has sufficient speed and less than 1 ms of delay, but the information change cycle and data processing time of SDN are factors that greatly affect the delay. Especially, in an emergency of self-driving environment linked to an ITS(Intelligent Traffic System) that requires low latency and high reliability, information should be transmitted and processed very quickly. That is a case in point where delay plays a very sensitive role. In this paper, we study the SDN architecture in emergencies during autonomous driving and conduct analysis through simulation of the correlation with the cell layer in which the vehicle should request relevant information according to the information flow. For simulation: As the Data Rate of 5G is high enough, we can assume the information for neighbor vehicle support to the car without errors. Furthermore, we assumed 5G small cells within 50 ~ 250 m in cell radius, and the maximum speed of the vehicle was considered as a 30km ~ 200 km/hour in order to examine the network architecture to minimize the delay.

• Journal of Intelligence and Information Systems
• /
• v.27 no.1
• /
• pp.83-102
• /
• 2021

The government recently announced various policies for developing big-data and artificial intelligence fields to provide a great opportunity to the public with respect to disclosure of high-quality data within public institutions. KSURE(Korea Trade Insurance Corporation) is a major public institution for financial policy in Korea, and thus the company is strongly committed to backing export companies with various systems. Nevertheless, there are still fewer cases of realized business model based on big-data analyses. In this situation, this paper aims to develop a new business model which can be applied to an ex-ante prediction for the likelihood of the insurance accident of credit guarantee. We utilize internal data from KSURE which supports export companies in Korea and apply machine learning models. Then, we conduct performance comparison among the predictive models including Logistic Regression, Random Forest, XGBoost, LightGBM, and DNN(Deep Neural Network). For decades, many researchers have tried to find better models which can help to predict bankruptcy since the ex-ante prediction is crucial for corporate managers, investors, creditors, and other stakeholders. The development of the prediction for financial distress or bankruptcy was originated from Smith(1930), Fitzpatrick(1932), or Merwin(1942). One of the most famous models is the Altman's Z-score model(Altman, 1968) which was based on the multiple discriminant analysis. This model is widely used in both research and practice by this time. The author suggests the score model that utilizes five key financial ratios to predict the probability of bankruptcy in the next two years. Ohlson(1980) introduces logit model to complement some limitations of previous models. Furthermore, Elmer and Borowski(1988) develop and examine a rule-based, automated system which conducts the financial analysis of savings and loans. Since the 1980s, researchers in Korea have started to examine analyses on the prediction of financial distress or bankruptcy. Kim(1987) analyzes financial ratios and develops the prediction model. Also, Han et al.(1995, 1996, 1997, 2003, 2005, 2006) construct the prediction model using various techniques including artificial neural network. Yang(1996) introduces multiple discriminant analysis and logit model. Besides, Kim and Kim(2001) utilize artificial neural network techniques for ex-ante prediction of insolvent enterprises. After that, many scholars have been trying to predict financial distress or bankruptcy more precisely based on diverse models such as Random Forest or SVM. One major distinction of our research from the previous research is that we focus on examining the predicted probability of default for each sample case, not only on investigating the classification accuracy of each model for the entire sample. Most predictive models in this paper show that the level of the accuracy of classification is about 70% based on the entire sample. To be specific, LightGBM model shows the highest accuracy of 71.1% and Logit model indicates the lowest accuracy of 69%. However, we confirm that there are open to multiple interpretations. In the context of the business, we have to put more emphasis on efforts to minimize type 2 error which causes more harmful operating losses for the guaranty company. Thus, we also compare the classification accuracy by splitting predicted probability of the default into ten equal intervals. When we examine the classification accuracy for each interval, Logit model has the highest accuracy of 100% for 0~10% of the predicted probability of the default, however, Logit model has a relatively lower accuracy of 61.5% for 90~100% of the predicted probability of the default. On the other hand, Random Forest, XGBoost, LightGBM, and DNN indicate more desirable results since they indicate a higher level of accuracy for both 0~10% and 90~100% of the predicted probability of the default but have a lower level of accuracy around 50% of the predicted probability of the default. When it comes to the distribution of samples for each predicted probability of the default, both LightGBM and XGBoost models have a relatively large number of samples for both 0~10% and 90~100% of the predicted probability of the default. Although Random Forest model has an advantage with regard to the perspective of classification accuracy with small number of cases, LightGBM or XGBoost could become a more desirable model since they classify large number of cases into the two extreme intervals of the predicted probability of the default, even allowing for their relatively low classification accuracy. Considering the importance of type 2 error and total prediction accuracy, XGBoost and DNN show superior performance. Next, Random Forest and LightGBM show good results, but logistic regression shows the worst performance. However, each predictive model has a comparative advantage in terms of various evaluation standards. For instance, Random Forest model shows almost 100% accuracy for samples which are expected to have a high level of the probability of default. Collectively, we can construct more comprehensive ensemble models which contain multiple classification machine learning models and conduct majority voting for maximizing its overall performance.