• Journal of Intelligence and Information Systems
• /
• v.7 no.2
• /
• pp.51-63
• /
• 2001

In supervised machine learning, an induction algorithm, which is able to extract rules from data with learning capability, provides a useful tool for data mining. Practical induction algorithms are known to degrade in prediction accuracy and generate complex rules unnecessarily when trained on data containing superfluous features. Thus it needs feature subset selection for better performance of them. In feature subset selection on the induction algorithm, wrapper method is repeatedly run it on the dataset using various feature subsets. But it is impractical to search the whole space exhaustively unless the features are small. This study proposes a heuristic method that uses sensitivity analysis of neural networks to the wrapper method for generating rules with higher possible accuracy. First it gives priority to all features using sensitivity analysis of neural networks. And it uses the wrapper method that searches the ordered feature space. In experiments to three datasets, we show that the suggested method is capable of selecting a feature subset that improves the performance of the induction algorithm within certain iteration.

• The Journal of the Korea Contents Association
• /
• v.9 no.10
• /
• pp.51-58
• /
• 2009

Reinforcement learning is a sub area of machine learning concerned with how an agent ought to take actions in an environment so as to maximize some notion of long-term reward. In the case of multi-agent, especially, which state space and action space gets very enormous in compared to single agent, so it needs to take most effective measure available select the action strategy for effective reinforcement learning. This paper proposes a multi-agent reinforcement learning model based on fuzzy inference system in order to improve learning collect speed and select an effective action in multi-agent. This paper verifies an effective action select strategy through evaluation tests based on Robocup Keepaway which is one of useful test-beds for multi-agent. Our proposed model can apply to evaluate efficiency of the various intelligent multi-agents and also can apply to strategy and tactics of robot soccer system.

• Journal of Intelligence and Information Systems
• /
• v.15 no.3
• /
• pp.1-15
• /
• 2009

In a classification problem, data imbalance occurs when the number of instances in one class greatly outnumbers the number of instances in the other class. Such data sets often cause a default classifier to be built due to skewed boundary and thus the reduction in the classification accuracy of such a classifier. This paper proposes a Geometric Mean-based Boosting (GM-Boost) to resolve the problem of data imbalance. Since GM-Boost introduces the notion of geometric mean, it can perform learning process considering both majority and minority sides, and reinforce the learning on misclassified data. An empirical study with bankruptcy prediction on Korea companies shows that GM-Boost has the higher classification accuracy than previous methods including Under-sampling, Over-Sampling, and AdaBoost, used in imbalanced data and robust learning performance regardless of the degree of data imbalance.

• Journal of Intelligence and Information Systems
• /
• v.15 no.4
• /
• pp.141-158
• /
• 2009

The requirements of knowledge customization have increased as information resources have become more various and the numbers of the resources are increased. Even if the method for collecting the information has improved like Really Simple Syndication (RSS), information users are still struggling for extracting and customizing the required information through the Web. To reduce the burden, we offer the dynamic knowledge customization framework by using ontology-based CBR. The framework consisting of three phases is comprised of the conversion phase of web information as a machine-accessible case, the extraction phase to find a case appropriate for information users' requirements, and the case customization phase to create knowledge depending on information user's requirements. Newly, the dynamic and intensity-based similarity is adopted to support timely dynamic change of users' requirements. The framework has adopted to create traveler's knowledge to the level users wanted.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2018.05a
• /
• pp.332-335
• /
• 2018

In recent years, artificial intelligence using big data has become a big issue in IT. Various studies are being conducted on services or technologies to effectively handle big data. The educational field, there is big data about students, but it is only a simple process to collect, lookup and store such data. In the future, it makes extensive use of artificial intelligence, machine learning, and statistical analysis to find meaningful rules, patterns, and relationships in the big data of the educational field, and to produce intelligent and useful data for the actual students. Accordingly, this study aims to design a program to predict the career of students using a decision tree algorithm based on the data from the student's classroom observations. Through a career prediction program, it is believed to be helpful to present application paths to students ' counseling and to also provide classroom behavior and direction based on the desired courses.

• KIPS Transactions on Computer and Communication Systems
• /
• v.5 no.1
• /
• pp.17-26
• /
• 2016

IoT means things space networks that form the intelligent relationship such as sensing, networking, information processing about human being, things and service without explicit mutual cooperation of human being. Lately many IoT groups such as AllSeen Alliance, OIC launched a platform for IoT. Self-adaptive is aimed at implementation without the need for decisions of human being during the operation, so that the machine can respond to changes in its own determination. There is a need to apply the concept of self-adaptive to existing IoT and IoT platform. So In this paper, We look for trends of existing IoT, IoT platform and comparisons by applying a self-adaptive concept to IoT, IoT platform. In addition as an example of this paper, we suggest lacking self-adaptive elements to OIC.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1612-1616
• /
• 2005

In this paper, a Home Mess-Cleanup Robot(HMR), which has a practical function of the automatic mess-cleanup, is developed. The vacuum-cleaner had made the burden of house chore lighten but the operation labour of a vacuum-cleaner had been so severe. Recently, the cleaning robot was producted to perfectly solve the cleaning labour of a house but it also was not successful because it still had a problem of mess-cleaning, which was the clean-up of big trash and the arrangement of newspapers, clothes, etc. The cleaning robot is to just vacuum dust and small trash but has no function to arrange and take away before the automatic vacuum-cleaning. For this reason, the market for the cleaning robot is not yet built up. So, we need a design method and technological algorithm of new automatic machine to solve the problem of mess-cleanup in house. It needs functions of agile automatic navigation, novel manipulation system for mess-cleanup. The automatic navigation system has to be controlled for the full scanning of living room, to recognize the absolute position and orientation of the self, the precise tracking of the desired path, and to distinguish the mess object to clean-up from obstacle object to just avoid. The manipulator, which is not needed in the vacuum-cleaning robot, must have the functions, how to distinguish big trash to clean from mess objects to arrange, how to grasp in according to the form of mess objects, how to move to the destination in according to mess objects and arrange them. We use the RFID system to solve the problems in this paper and propose the reading algorithm of RFID tags installed in indoor objects and environments. Then, it should be an intelligent system so that the mess cleaning task can be autonomously performed in a wide variety of situations and environments. It needs to also has the entertainment functions for the good communication between the human and HMR. Finally, the good performance of the designed HMR is confirmed through the results of the mess clean-up and arrangement.

• Journal of the Korean Data and Information Science Society
• /
• v.24 no.2
• /
• pp.223-233
• /
• 2013

At local emerging stock markets such as Korea, Hong Kong, Singapore and Taiwan, foreign investors (FI) are recognized as important investment community due to the globalization and deregulation of financial markets. Therefore, it is required to monitor the behavior of FI against a sudden enormous selling stocks for the concerned local governments or private and institutional investors. The main aim of this study is to propose an early warning system (EWS) which purposes issuing a warning signal against the possible massive selling stocks of FI at the market. For this, we suggest machine learning algorithm which predicts the behavior of FI by forecasting future conditions. This study is empirically done for the Korean stock market.

• Journal of the Korean Institute of Telematics and Electronics
• /
• v.26 no.2
• /
• pp.76-86
• /
• 1989

Many intelligent robots that are equipped with special tools and sensors re currently used in assembly line. As automatic manufacturing systems including such robots become advanced and complicated, there are increasing needs for the development of the sophisticated programming systems which can control several robots and other manufacutring equipments in workcell at a time. In this paper a programming language, ARL (Assembly Robot Language), is proposed and developed, which can control the manufacturing devices as well as robots in workcell. It has not only all the common features of modern textual robot language but also debugging facilities. In this language system machine dependecy is minimized by using dedicated processes and a shared memory for communication between processes. Extensibility and adaptability of the programming system is increased by using such a technique against the changes of workcel environment.

• Journal of Intelligence and Information Systems
• /
• v.14 no.4
• /
• pp.179-200
• /
• 2008

We develop a feature selection method that can improve both the efficiency and the effectiveness of classification technique. In this research, we employ case-based reasoning as a classification technique. Basically, this research integrates the two existing feature selection methods, i.e., the univariate analysis and the LVF algorithm. First, we sift some predictive features from the whole set of features using the univariate analysis. Then, we generate all possible subsets of features from these predictive features and measure the inconsistency rate of each subset using the LVF algorithm. Finally, the subset having the lowest inconsistency rate is selected as the best subset of features. We measure the performances of our feature selection method using the data obtained from UCI Machine Learning Repository, and compare them with those of existing methods. The number of selected features and the accuracy of our feature selection method are so satisfactory that the improvements both in efficiency and effectiveness are achieved.