• Journal of the Korean Association of Geographic Information Studies
• /
• v.21 no.3
• /
• pp.205-213
• /
• 2018

Hyperspectral image is more increasing spectral resolution that Multi-spectral image. Because of that, each pixel of the hyperspectral image includes much more information and it is considered the most appropriate technic for land cover classification. but recent research of hyperspectral image is stayed land cover classification of general level. therefore we classified land cover of detail level using ED, SAM, SSS method and made Decision Tree from result of that. As a result, the overall accuracy of general level was improved by 1.68% and the overall accuracy of detail level was improved by 5.56%.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.10
• /
• pp.59-68
• /
• 1998

The goal of data mining is to develop the automatic and intelligent tools and technologies that can find useful knowledge from databases. To meet this goal, we propose an efficient data mining algorithm based on the fuzzy decision tree. The proposed method combines comprehensibility of decision tree such as ID3 and C4.5 and representation power of fuzzy set theory. So, it can generate simple and comprehensive rules describing data. The proposed algorithm consists of two stages: the first stage generates the fuzzy membership functions using histogram analysis, and the second stage constructs a fuzzy decision tree using the fuzzy membership functions. From the testing of the proposed algorithm on the IRIS data and the Wisconsin Breast Cancer data, we found that the proposed method can generate a set of fuzzy rules from data efficiently.

• Journal of Korea Artificial Intelligence Association
• /
• v.1 no.2
• /
• pp.1-6
• /
• 2023

In this paper, we present a new method for classifying malicious URLs to reduce cases of learning difficulties due to unfamiliar and difficult terms related to information protection. This study plans to extract only visually distinguishable features within the URL structure and compare them through map learning algorithms, and to compare the contribution values of the best map learning algorithm methods to extract features that have the most impact on classifying malicious URLs. As research data, Kaggle used data that classified 7,046 malicious URLs and 7.046 normal URLs. As a result of the study, among the three supervised learning algorithms used (Decision Tree, Support Vector Machine, and Logistic Regression), the Decision Tree algorithm showed the best performance with 83% accuracy, 83.1% F1-score and 83.6% Recall values. It was confirmed that the contribution value of https is the highest among whether to use https, sub domain, and prefix and suffix, which can be visually distinguished through the feature contribution of Decision Tree. Although it has been difficult to learn unfamiliar and difficult terms so far, this study will be able to provide an intuitive judgment method without explanation of the terms and prove its usefulness in the field of malicious URL detection.

• Clinical and Experimental Emergency Medicine
• /
• v.5 no.4
• /
• pp.230-239
• /
• 2018

Objective The suicide rate in South Korea is very high and is expected to increase in coming years. Intoxication is the most common suicide attempt method as well as one of the common reason for presenting to an emergency medical center. We used decision tree modeling analysis to identify predictors of risk for suicide by intentional intoxication. Methods A single-center, retrospective study was conducted at our hospital using a 4-year registry of the institute from January 1, 2013 to December 31, 2016. Demographic factors, such as sex, age, intentionality, therapeutic adherence, alcohol consumption, smoking status, physical disease, cancer, psychiatric disease, and toxicological factors, such as type of intoxicant and poisoning severity score were collected. Candidate risk factors based on the decision tree were used to select variables for multiple logistic regression analysis. Results In total, 4,023 patients with intoxication were enrolled as study participants, with 2,247 (55.9%) identified as cases of intentional intoxication. Reported annual percentages of intentional intoxication among patients were 628/937 (67.0%), 608/1,082 (56.2%), 536/1,017 (52.7), 475/987 (48.1%) from 2013 to 2016. Significant predictors identified based on decision tree analysis were alcohol consumption, old age, psychiatric disease, smoking, and male sex; those identified based on multiple regression analysis were alcohol consumption, smoking, male sex, psychiatric disease, old age, poor therapeutic adherence, and physical disease. Conclusion We identified important predictors of suicide risk by intentional intoxication. A specific and realistic approach to analysis using the decision tree modeling technique is an effective method to determine those groups at risk of suicide by intentional intoxication.

• Journal of the Ergonomics Society of Korea
• /
• v.25 no.2
• /
• pp.161-172
• /
• 2006

This study aims to compare and contrast the Kansei modeling methods for building a luxuriousness model that people feel about appearance of mobile phones. For the evaluation based on Kansei engineering approaches, 15 participants were employed to evaluate 18 mobile phones using a questionnaire. The results of evaluation were analyzed to build luxuriousness models through quantification I method, neural network, and decision tree method, respectively. The performance of Kansei modeling methods was compared and contrasted in terms of accuracy and predictability. The result of comparison of modeling methods indicated that model accuracy and predictability was closely related to the number of variables and data size. It was also revealed that quantification I method was the best in terms of model accuracy while decision tree method was the best modeling method with small variance in terms of predictability. However, it was empirically found that quantification I method showed extremely unstable predictability with small number of data. Consequently, it is expected that the research findings of this study might be utilized as a guideline for selecting proper Kansei modeling method.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.31 no.4
• /
• pp.49-58
• /
• 2008

This paper proposes a methodology to measure the risk level in real-time for Business Activity Monitoring (BAM). A decision-tree methodology was employed to analyze the effect of process attributes on the result of the process execution. In the course of process execution, the level of risk is monitored in real-time, and an early warning can be issued depending on the change of the risk level. An algorithm for estimating the risk of ongoing processes in real-time was formulated. Comparison experiments were conducted to demonstrate the effectiveness of our method. The proposed method detects the risks of business processes more precisely and even earlier than existing approaches.

• Journal of KIISE:Databases
• /
• v.33 no.5
• /
• pp.463-475
• /
• 2006

In applications such as a traffic management system, analysis using a spatial hierarchy of a spatial data warehouse and a simple aggregation is required. Over the past few years, several studies have been made on solution using a spatial index. Many studies have focused on using extended R-tree. But, because it just provides either the current aggregation or the total aggregation, decision support of traffic policy required historical analysis can not be provided. This paper proposes hybrid index based on extended aR-tree for the spatio-temporal aggregation. The proposed method supports a spatial hierarchy and the current aggregation by the R-tree. The sorted hash table using the time structure of the extended aR-tree provides a temporal hierarchy and a historical aggregation. Therefore, the proposed method supports an efficient decision support with spatio-temporal analysis and is Possible currently traffic analysis and determination of a traffic policy with historical analysis.

• Journal of Korea Society of Digital Industry and Information Management
• /
• v.19 no.4
• /
• pp.1-11
• /
• 2023

Clustering analysis is used in various fields including customer segmentation and clustering methods such as k-means are actively applied in the credit card customer segmentation. In this paper, we summarized the input features selection method of k-means clustering for the case of the credit card customer segmentation problem, and evaluated its feasibility through the analysis results. By using the label values of k-means clustering results as target features of a decision tree classification, we composed a method for prioritizing input features using the information gain of the branch. It is not easy to determine effectiveness with the clustering effectiveness index, but in the case of the CH index, cluster effectiveness is improved evidently in the method presented in this paper compared to the case of randomly determining priorities. The suggested method can be used for effectiveness of actively used clustering analysis including k-means method.

• Journal of Information Technology Applications and Management
• /
• v.15 no.2
• /
• pp.51-65
• /
• 2008

The purpose of this paper is introducing WiFi based EKF(Extended Kalman Filter) method for indoor positioning. The advantages of our EKF method include: 1) Any special equipment dedicated for positioning is not required. 2) implementation of EKF does not require off-line phase of fingerprinting methods. 3) The EKF effectively minimizes squared deviation of the trilateration method. In order to experimentally prove the advantages of our method, we implemented indoor positioning systems making use of the K-NN(K Nearest Neighbors), Bayesian, decision tree, trilateration, and our EKF methods. Our experimental results show that the average-errors of K-NN, Bayesian and decision tree methods are all close to 2.4 meters whereas the average errors of trilateration and EKF are 4.07 meters and 3.528 meters, respectively. That is, the accuracy of our EKF is a bit inferior to those of fingerprinting methods. Even so, our EKF is accurate enough to be used for practical indoor LBS systems. Moreover, our EKF is easier to implement than fingerprinting methods because it does not require off-line phase.

• Proceedings of the KIEE Conference
• /
• 1998.07g
• /
• pp.2330-2332
• /
• 1998

This paper suggests a method to recognize the various defect patterns of cold mill strip using binary decision tree automatically constructed by genetic algorithm. The genetic algorithm and K-means algorithm were used to select a subset of the suitable features at each node in binary decision tree. The feature subset with maximum fitness is chosen and the patterns are classified into two classes by a linear decision boundary. This process was repeated at each node until all the patterns are classified into individual classes. The final recognizer is accomplished by neural network learning of a set of standard patterns at each node. Binary decision tree classifier was applied to the recognition of the defect patterns of cold mill strip and the experimental results were given to demonstrate the usefulness of the proposed scheme.