• The Journal of the Korea Contents Association
• /
• v.18 no.8
• /
• pp.525-535
• /
• 2018

If the frequency of a particular class is excessively higher than the frequency of other classes in the classification problem, data imbalance problems occur, which make machine learning distorted. Corporate bankruptcy prediction often suffers from data imbalance problems since the ratio of insolvent companies is generally very low, whereas the ratio of solvent companies is very high. To mitigate these problems, it is required to apply a proper sampling technique. Until now, oversampling techniques which adjust the class distribution of a data set by sampling minor class with replacement have popularly been used. However, they are a risk of overfitting. Under this background, this study proposes ROSE(Random Over Sampling Examples) technique which is proposed by Menardi and Torelli in 2014 for the effective corporate bankruptcy prediction. The ROSE technique creates new learning samples by synthesizing the samples for learning, so it leads to better prediction accuracy of the classifiers while avoiding the risk of overfitting. Specifically, our study proposes to combine the ROSE method with SVM(support vector machine), which is known as the best binary classifier. We applied the proposed method to a real-world bankruptcy prediction case of a Korean major bank, and compared its performance with other sampling techniques. Experimental results showed that ROSE contributed to the improvement of the prediction accuracy of SVM in bankruptcy prediction compared to other techniques, with statistical significance. These results shed a light on the fact that ROSE can be a good alternative for resolving data imbalance problems of the prediction problems in social science area other than bankruptcy prediction.

• Journal of Sasang Constitutional Medicine
• /
• v.19 no.3
• /
• pp.51-61
• /
• 2007

1. Objectives The Face is an important standard for the classification of Sasang Constitutions. We are developing 3D Automatic Face Recognition Apparatus(3D-AFRA) to analyse the facial characteristics. This apparatus show us 3D image and data of man's face and measure facial figure data. So We should examine the Measurement of Facial Figure data error of 3D Automatic Face Recognition Apparatus(3D-AFRA) in Software Error Analysis. 2. Methods We scanned face status by using 3D Automatic Face Recognition Apparatus(3D-AFRA). And we measured lengths Between Facial Definition Parameters of facial figure data by Facial Measurement program. 2.1 Repeatability test We measured lengths Between Facial Definition Parameters of facial figure data restored by 3D-AFRA by Facial Measurement program 10 times. Then we compared 10 results each other for repeatability test. 2.2 Measurement error test We measured lengths Between Facial Definition Parameters of facial figure data by two different measurement program that are Facial Measurement program and Rapidform2006. At measuring lengths Between Facial Definition Parameters, we uses two measurement way. The one is straight line measurement, the other is curved line measurement. Then we compared results measured by Facial Measurement program with results measured by Rapidform2006. 3. Results and Conclusions In repeatability test, standard deviation of results is 0.084-0.450mm. And in straight line measurement error test, the average error 0.0582mm, and the maximum error was 0.28mm. In curved line measurement error test, the average error 0.413mm, and the maximum error was 1.53mm. In conclusion, we assessed that the accuracy and repeatability of Facial Measurement program is considerably good. From now on we complement accuracy of 3D-AFRA in Hardware and Software.

• Journal of the Korean Association of Oral and Maxillofacial Surgeons
• /
• v.26 no.5
• /
• pp.462-469
• /
• 2000

Positron Emission Tomography(PET) is a new diagnostic method that can create functional images of the distribution of positron emitting radionuclides, which when administered intravenously in the body, makes possible anatomical and functional analysis by quantity of biochemical and physiological process. After genetic and biochemical changes in initial stage, malignant tumor undergoes functional changes before undergoing anatomical changes. So, early diagnosis of malignant tumors by functional analysis with PET can be achieved, replacing traditional anatomical analysis, such as computed tomography(CT) and magnetic resonance image(MRI), etc. Similarly, PET can identify malignant tumor without confusion with scar and fibrosis in follow up check. In the Korea Cancer Center Hospital(KCCH) from October 1997 to September 1999, clinical study was performed in 79 cases that underwent 89 times PET evaluation with [18F]-Fluorodeoxyglucose for diagnosis of oral and maxillofacial tumors, and the data was analysed by Bayesian $2{\times}2$ Classification Table. The results were as follows : Evaluation for initial diagnosis with FDG-PET (P<0.005) 1. Agreement rate or accuracy rate is 88.9%. 2. Sensitivity is 95.2%, and specificity 66.7%. 3. Positive predictive rate is 90.9%, and negative predictive rate 80.0%. 4. In consideration of tumor stage, diagnostic rate in less than stage II was 90% and in greater than stage III 100%. 5. In consideration of tumor size, diagnostic rate in less than T2 was 92.3% and in greater than T3 100%. After primary treatment, evaluation for follow up check with FDG-PET (P < 0.001) 1. Agreement rate or accuracy rate is 85.4%. 2. Sensitivity is 87.5%, and specificity 82.4%. 3. Positive predictive rate is 87.5%, and negative predictive rate 82.4%. 4. In 24 recurred cases, 6 had distant metastasis, and 5 of them were diagnosed with FDG-PET, resulting in diagnostic rate of FDG-PET of 83.3%. From the above results, Positron Emission Tomography with [18F]- Fluorodeoxyglucose appears to be more sensitive and accurate for detecting the presence of oral and maxillofacial tumors, and has various clinical applications such as early diagnosis of tumor in initial and follow up check and detection of distant metastasis.

• Journal of Internet Computing and Services
• /
• v.19 no.5
• /
• pp.33-41
• /
• 2018

Recently, Internet technology has developed, various programs are being created and therefore various codes are being made through many authors. On this aspect, some author deceive a program or code written by other particular author as they make it themselves and use other writers' code indiscriminately, or not indicating the exact code which has been used. Due to this makes it more and more difficult to protect the code. In this paper, we propose author identification framework using Authorship Analysis theory and Natural Language Processing(NLP) based on Convolutional Neural Network(CNN). We apply Authorship Analysis theory to extract features for author identification in the source code, and combine them with the features being used text mining to perform author identification using machine learning. In addition, applying CNN based natural language processing method to source code for code author classification. Therefore, we propose a framework for the identification of authors using the Authorship Analysis theory and the CNN. In order to identify the author, we need special features for identifying the authors only, and the NLP method based on the CNN is able to apply language with a special system such as source code and identify the author. identification accuracy based on Authorship Analysis theory is 95.1% and identification accuracy applied to CNN is 98%.

• Journal of Internet Computing and Services
• /
• v.18 no.1
• /
• pp.65-75
• /
• 2017

The Detection Model is the model to find the result of a certain purpose using artificial intelligent, data mining, intelligent algorithms In Cyber Security, it usually uses to detect intrusion, malwares, cyber incident, and attacks etc. There are an amount of unlabeled data that are collected in a real environment such as security data. Since the most of data are not defined the class labels, it is difficult to know type of data. Therefore, the label determination process is required to detect and analysis with accuracy. In this paper, we proposed a KDFL(K-means and D-S Fusion based Labeling) method using D-S inference and k-means(unsupervised) algorithms to decide label of data records by fusion, and a detection model architecture using a proposed labeling method. A proposed method has shown better performance on detection rate, accuracy, F1-measure index than other methods. In addition, since it has shown the improved results in error rate, we have verified good performance of our proposed method.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.10
• /
• pp.8-15
• /
• 2020

A forecasting method using deep learning does not have consistent results due to the differences in the characteristics of the dataset, even though they have the same forecasting models and parameters. For example, the forecasting model X optimized with dataset A would not produce the optimized result with another dataset B. The forecasting model with the characteristics of the dataset needs to be optimized to increase the accuracy of the forecasting model. Therefore, this paper proposes novel optimization steps for outlier removal, dataset classification, and a CNN-LSTM-based hyperparameter tuning process to forecast the daily power usage of a university campus based on the hourly interval. The proposing model produces high forecasting accuracy with a 2% of MAPE with a single power input variable. The proposing model can be used in EMS to suggest improved strategies to users and consequently to improve the power efficiency.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.27 no.5
• /
• pp.619-628
• /
• 2009

LiDAR has capability of rapid data acquisition and provides useful information for reconstructing surface of the Earth. However, Extracting information from LiDAR data is not easy task because LiDAR data consist of irregularly distributed point clouds of 3D coordinates and lack of semantic and visual information. This thesis proposed methods for automatic extraction of buildings and 3D detail modeling using airborne LiDAR data. As for preprocessing, noise and unnecessary data were removed by iterative surface fitting and then classification of ground and non-ground data was performed by analyzing histogram. Footprints of the buildings were extracted by tracing points on the building boundaries. The refined footprints were obtained by regularization based on the building hypothesis. The accuracy of building footprints were evaluated by comparing with 1:1,000 digital vector maps. The horizontal RMSE was 0.56m for test areas. Finally, a method of 3D modeling of roof superstructure was developed. Statistical and geometric information of the LiDAR data on building roof were analyzed to segment data and to determine roof shape. The superstructures on the roof were modeled by 3D analytical functions that were derived by least square method. The accuracy of the 3D modeling was estimated using simulation data. The RMSEs were 0.91m, 1.43m, 1.85m and 1.97m for flat, sloped, arch and dome shapes, respectively. The methods developed in study show that the automation of 3D building modeling process was effectively performed.

• Korean Journal of Food Science and Technology
• /
• v.49 no.5
• /
• pp.469-473
• /
• 2017

The objective of this study was to develop methods for the discrimination of new and stale rice by using an electronic eye. To develop the discriminant, 107 rice samples produced in the years 2015 and 2016 were investigated. After the rice was treated with guaiacol, oxydol, and p-phenylenediamine reagents, an electronic eye was applied to discriminate between newly harvested rice and rice stored for 1 year. Out of the 4,096 color codes of the electronic eye, 31 color codes were identified for the discrimination of newly harvested rice and rice stored for 1 year. The classification ratio of newly harvested rice and rice stored for 1 year was 100% and the discrimination accuracy for unknown samples was 100%. In a total of 150 mixtures of new rice and stale rice, the discrimination accuracy was between 16.7 and 95.6%, depending on the mixing ratio. This capability of the electronic eye will be useful as a tool for discriminating the production year of rice.

• Journal of the Korean Academy of Child and Adolescent Psychiatry
• /
• v.23 no.2
• /
• pp.69-75
• /
• 2012

Objectives : The current study was conducted in order to investigate teachers' recognition of school bullying using a nationwide database of adolescents in middle and high school in Korea. Methods : Students in the 7th to 12th grades at 23 secondary schools participated in the current study during the fall of 2009. Subjects completed the self-report form of the Adolescent Mental Health and Problem Behavior Screening Questionnaire-II (AMPQ-II) and Symptom Checklist-90 Revised (SCL-90-R). In addition, relevant teachers used the teachers' rating scale of the AMPQ-II to report their students' status. Differences in the number of bullied students between teachers' recognition and students' report were explored. Results : A total of 2270 subjects provided relevant responses to the questionnaire. While the one-month prevalence of victimization according to students' self-reports was 28.9%, the recognized prevalence by teachers was only 10.6%. For prediction of the presence of school bullying according to students' self reports on the AMPQ-II, item 7 of the teachers' report on the AMPQ-II showed a sensitivity of 16%, a specificity of 92%, a positive predictability of 44%, a negative predictability of 72%, a false positive rate of 8%, a false negative rate of 84%, and an accuracy of 69%, respectively. No significant differences in subscores of students' self reports of the AMPQ-II and SCL-90-R were observed between bullied students who were recognized by teachers and those who were not recognized. In stepwise discriminant analysis, classification of teachers' item 2 and item 7 on the AMPQ-II with respect to school bullying according to students' reports showed an accuracy of 63.4%. Using this model, 75.2% of non-victimized subjects were classified correctly, while only 35.2% of victimized subjects were classified correctly. Conclusion : Despite the high prevalence in Korea, teachers' recognition of school violence among their students remains low. Pre-professional and continuing education to improve teachers' understanding of school bullying and knowledge of effective classroom-based prevention activities should be encouraged.

• Journal of Information Technology and Architecture
• /
• v.10 no.1
• /
• pp.101-111
• /
• 2013

Internet traffic volume has been increasing rapidly due to popularization of various smart devices and Internet development. In particular, HTTP-based traffic volume of smart devices is increasing rapidly in addition to desktop traffic volume. The increased mobile traffic can cause serious problems such as network overload, web security, and QoS. In order to solve these problems of the Internet overload and security, it is necessary to accurately detect applications. Traditionally, well-known port based method is utilized in traffic classification. However, this method shows low accuracy since P2P applications exploit a TCP/80 port, which is used for the HTTP protocol; to avoid firewall or IDS. Signature-based method is proposed to solve the lower accuracy problem. This method shows higher analysis rate but it has overhead of signature generation. Also, previous signature-based study only analyzes applications in HTTP protocol-level not application-level. That is, it is difficult to identify application name. Therefore, previous study only performs protocol-level analysis. In this paper, we propose a signature generation method to classify HTTP-based traffics in application-level using the characteristics of typical semi HTTP header. By applying our proposed method to campus network traffic, we validate feasibility of our method.