• Journal of Intelligence and Information Systems
• /
• v.18 no.1
• /
• pp.125-141
• /
• 2012

These days, the malicious attacks and hacks on the networked systems are dramatically increasing, and the patterns of them are changing rapidly. Consequently, it becomes more important to appropriately handle these malicious attacks and hacks, and there exist sufficient interests and demand in effective network security systems just like intrusion detection systems. Intrusion detection systems are the network security systems for detecting, identifying and responding to unauthorized or abnormal activities appropriately. Conventional intrusion detection systems have generally been designed using the experts' implicit knowledge on the network intrusions or the hackers' abnormal behaviors. However, they cannot handle new or unknown patterns of the network attacks, although they perform very well under the normal situation. As a result, recent studies on intrusion detection systems use artificial intelligence techniques, which can proactively respond to the unknown threats. For a long time, researchers have adopted and tested various kinds of artificial intelligence techniques such as artificial neural networks, decision trees, and support vector machines to detect intrusions on the network. However, most of them have just applied these techniques singularly, even though combining the techniques may lead to better detection. With this reason, we propose a new integrated model for intrusion detection. Our model is designed to combine prediction results of four different binary classification models-logistic regression (LOGIT), decision trees (DT), artificial neural networks (ANN), and support vector machines (SVM), which may be complementary to each other. As a tool for finding optimal combining weights, genetic algorithms (GA) are used. Our proposed model is designed to be built in two steps. At the first step, the optimal integration model whose prediction error (i.e. erroneous classification rate) is the least is generated. After that, in the second step, it explores the optimal classification threshold for determining intrusions, which minimizes the total misclassification cost. To calculate the total misclassification cost of intrusion detection system, we need to understand its asymmetric error cost scheme. Generally, there are two common forms of errors in intrusion detection. The first error type is the False-Positive Error (FPE). In the case of FPE, the wrong judgment on it may result in the unnecessary fixation. The second error type is the False-Negative Error (FNE) that mainly misjudges the malware of the program as normal. Compared to FPE, FNE is more fatal. Thus, total misclassification cost is more affected by FNE rather than FPE. To validate the practical applicability of our model, we applied it to the real-world dataset for network intrusion detection. The experimental dataset was collected from the IDS sensor of an official institution in Korea from January to June 2010. We collected 15,000 log data in total, and selected 10,000 samples from them by using random sampling method. Also, we compared the results from our model with the results from single techniques to confirm the superiority of the proposed model. LOGIT and DT was experimented using PASW Statistics v18.0, and ANN was experimented using Neuroshell R4.0. For SVM, LIBSVM v2.90-a freeware for training SVM classifier-was used. Empirical results showed that our proposed model based on GA outperformed all the other comparative models in detecting network intrusions from the accuracy perspective. They also showed that the proposed model outperformed all the other comparative models in the total misclassification cost perspective. Consequently, it is expected that our study may contribute to build cost-effective intelligent intrusion detection systems.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.10
• /
• pp.1278-1285
• /
• 2018

For saving electricity bill, energy storage system(ESS) is being installed in factories, public building and commercial building with a Time-of-Use(TOU) tariff which consists of demand charge(KRW/kW) and energy charge(KRW/kWh). However, both of peak reduction and ESS special tariff are not considered in an analysis of initial cost payback period(ICPP) on ESS. Since it is difficult to reflect base rate by an amount of uncertain peak demand reduction during mid-peak and on-peak periods in the future days. Therefore, the ICPP on ESS can be increased. Based on this background, this paper presents the advanced analysis method for the ICPP on ESS. In the proposed algorithm, the representative days of monthly electricity consumption pattern for the amount of peak reduction can be found by the k­means clustering algorithm. Moreover, the total expected energy costs of representative days are minimized by optimal daily ESS operation considering both peak reduction and the special tariff through a mixed-integer linear programming(MILP). And then, the amount of peak reduction becomes a value that the sum of the expected energy costs for 12 months is maximum. The annual benefit cost is decided by the amount of annual peak reduction. Two simulation cases are considered in this study, which one only considers the special tariff and another considers both of the special tariff and amount of peak reduction. The ICPP in the proposed method is shortened by 18 months compared to the conventional method.

• Journal of Korean Public Health Nursing
• /
• v.31 no.1
• /
• pp.5-18
• /
• 2017

Purpose: This study was conducted to suggest a method for financial projection of health insurance expenditures that reflects future changes in demographic structure. Methods: Using data associated with the number of patients and health insurance cost per patient, generalized linear models (GLM) were fitted with demographic explanatory variables. Models were constructed separately for individual medical departments, types of medical service, and types of public health insurance. Goodness-of-fit of most of the applied GLM models was quite satisfactory. By combining estimates of frequency and severity from the constructed models and results of the population projection, total annual health insurance expenditures were projected through year 2060. Results: Expenditures for medical departments associated with diseases that are more frequent in elderly peoples are expected to increase steeply, leading to considerable increases in overall health insurance expenditures. The suggested method can contribute to improvement of the accuracy of financial projection. Conclusion: The overall demands for medical service, medical personnel, and relevant facilities in the future are expected to increase as the proportion of elderly people increases. Application of a more reasonable estimation method reflecting changes in demographic structure will help develop health policies relevant to above mentioned resources.

• Industrial Engineering and Management Systems
• /
• v.14 no.4
• /
• pp.372-378
• /
• 2015

In labor-intensive enterprise, such as garment factory, assembly line is widely used as a manufacturing process for reducing costs and production time. However, for the sake of the various working capacity of worker, idle or delay may happen and influence the rear processes. If these unforeseeable delay happened continuously, it may influence the whole manufacturing process and a model, which is called limited-cycle model with multiple periods (LCMwMP), is assumed to evaluate the influence risk. In order to minimize the risk, the assignment of the workers is focused on. In this paper, we deal with an assembly line as LCMwMP model when two kinds of workers exist, whose efficiency is assumed to two different groups. We consider an optimization problem for finding an assignment of workers to the line that minimizes total expected risk, which is exchanged to expected cost by reset model of LCMwMP. First, reset model as a simple model of LCMwMP is introduced. Then, some hypotheses of the rules of the optimal worker assignment are proposed and some numerical experiments are researched assuming the processing time as Erlang distribution. Finally, the other rules on other certain conditions are discussed.

• Journal of the military operations research society of Korea
• /
• v.4 no.1
• /
• pp.113-123
• /
• 1978

Consider a series system of two units, named 1 and 2, respectively. Two units are observed at the beginning of discrete time periods t=0,1,2, $cdots$ and classified as being in one of a countable number of states. Let (i, r) be a state of the system at time t, when the state of unit 1 is i and state of unit 2 is r at time t, Under some conditions, the opportunistic replacement policy that minimizes the expected total discounted cost or the average cost of maintenance is shown to be characterized by the control limits $i^{*}(r)$ (a function of r) and $r^{*}(i)$ (a function of i) : (a) in observed state (i, r), the optimal policy for unit 1 is to replace if $i{\ge}i^{*}(r)$ and no action otherwise; (b) in observed state (i, r), the optimal policy for unit 2 is to replace if $r{\ge}r^{*}(i)$ and no action otherwise. In addition, this paper also develops optimal policy in the finite time horizon case, where time horizon is fixed or a finite integer valued r.v. with known pmf.

• Industrial Engineering and Management Systems
• /
• v.4 no.2
• /
• pp.145-157
• /
• 2005

There are a lot of raw materials, work-in-processes and finished goods in manufacturing industry. Here, the less stock of materials and work-in-processes manufacturing industry has, the worse the rate of the production is. Inversely, the more manufacturing industry has, the more expensive the cost to support them is. Thus, it is important for us to balance them efficiently. In general, inventory problems are to decide appropriate times to produce goods and to determine appropriate quantities of goods. Therefore, inventory problems require as more useful information as possible. For example, there are demand, lead time, ordering point and so on. In this paper, we deal with an optimal ordering policy on both way substitutable two-commodity inventory control system. That is, there is a problem of how to allocate the produced two kinds of goods in a factory to m areas so as to minimize the total expected inventory cost. The demand of each area is probabilistic, and we adopt the exponential distribution as a probability density function of demand. Moreover, we provide numerical examples of the problem.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.6B
• /
• pp.623-629
• /
• 2009

In this research, the process of developing software products to users in transfer by considering the warranty period to determine the timing of the release period is a comparative study of models. For the results of demonstration, exponential software reliability model increases the warranty period, the higher the initial period, but shows almost a similar release. In contrast, the optimal release time of imperfect debugging software reliability model, lower the initial warranty period, but the pattern is expected to rise slightly larger. The proposed model, extreme value distribution model, pattern of the optimal release time gradually increase, have a form that can be drawn. These research results through, warranty period and release the software developers understand the relationship between the optimal time for software development by using advance information could do is feed.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.51 no.2
• /
• pp.163-170
• /
• 2015

Recently LED fish-attracting lamps have been more widely used in fisheries as low-cost and high-efficiency fishing gear, and development of long-life high-efficiency lamps is required through the design of LED packages to optimize heat resistance. This study developed an improved LED fish-attracting lamp with excellent heat performance, which was verified using a numerical model. Heat-resistance design factors such as the heat-radiation fin shape, PCB type, and LED chip count were investigated and optimized. Comparison with a commercial 180-W LED fishing lamp showed that the increase in initial temperature was 40% higher than that of the surrounding LED chip because of design errors in contact thermal resistance. The 250-W LED lamp developed in this study has a characteristic with thermal rising in linearly stable according to the heat source. In addition, luminance efficiency was improved by 20-65% by using flow-visualization simulation. A decrease of 45% in total power consumption with a fuel-cost reduction of over 55% can be expected when using these optimized heat release design factors.

• Proceedings of the Korean Institute Of Construction Engineering and Management
• /
• 2007.11a
• /
• pp.853-858
• /
• 2007

Construction industry is traditionally labor intensive. And construction site is also in densely populated environment. So accurate labor monitoring is one of important issues for site management. Labor cost shares large amount of part in total construction cost. Therefore, it is a major element in site management and control. In this article, we explained development of prototype system for labor monitoring based on RFID technology and results of field tests using the system. According to the results, the system is able to expected to improve effectiveness of Labor management in construction site

• Journal of Society of Preventive Korean Medicine
• /
• v.12 no.3
• /
• pp.21-34
• /
• 2008

Objectives : This paper analysed the alternative methods of calculating conversion factor for oriental medicine in the National Health Insurance and estimated the conversion factor(reimbursing price level) of the oriental medical services, based on health insurance claims data and macro economic data. Methods : Comparing cost accounting method, SGR model, and index model to estimate conversion factor in the national health insurance, six empirical models were derived depending on the scope of revenue considered in financial indicators. Classifications of data and sources used in the analysis were identified as officially released by the government. Results and Conclusion : Cost accounting analysis and SGR model showed a two digit decrease in the physician fee schedule of oriental medical services in the national health insurance, while index model indicated a positive increase in the fee reimbursed. As expected, SGR model measured an overall trend of health expenditures rather than an individual financial status of medical institutions, and index model properly estimated the level of payments to oriental medical doctors. Upon a declining share of health expenditures on oriental medicine, a global budget system fixed to a flat rate of total budget could be an opportunity as well as a challenge.