• Journal of the Korea Institute of Military Science and Technology
• /
• v.24 no.3
• /
• pp.281-292
• /
• 2021

With the development of deep learning technology, researchers and technicians keep attempting to apply deep learning in various industrial and academic fields, including the defense. Most of these attempts assume that the data are balanced. In reality, since lots of the data are imbalanced, the classifier is not properly built and the model's performance can be low. Therefore, this study proposes cost-sensitive learning as a solution to the imbalance data problem of image classification in the defense field. In the proposed model, cost-sensitive learning is a method of giving a high weight on the cost function of a minority class. The results of cost-sensitive based model shows the test F1-score is higher when cost-sensitive learning is applied than general learning's through 160 experiments using submarine/non-submarine dataset and warship/non-warship dataset. Furthermore, statistical tests are conducted and the results are shown significantly.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10a
• /
• pp.107-110
• /
• 1996

Continuing advances in the formulation and solution of risk-sensitive control problems have reached a point at which this topic is becoming one of the more intriguing modern paradigms of feedback thought. Despite a prevailing atmosphere of close scrutiny of theoretical studies, the risk-sensitive body of knowledge is growing. Moreover, from the point of view of applications, the detailed properties of risk-sensitive design are only now beginning to be worked out. Accordingly, the time seems to be right for a survey of the historical underpinnings of the subject. This paper addresses the beginnings and the evolution, over the first quarter-century or so, and points out the close relationship of the topic with the notion of optimal cost cumulates, in particular the cost variance. It is to be expected that, in due course, some duality will appear between these notions and those in estimation and filtering. The purpose of this document is to help to lay a framework for that eventuality.

• Communications for Statistical Applications and Methods
• /
• v.11 no.1
• /
• pp.21-30
• /
• 2004

Often the accuracy is not adequate as a performance measure of classifiers when costs are different for different prediction errors. ROC and cost graphs can be used in such case to compare and identify cost-sensitive classifiers. We extend ROC and cost graphs so that they can be used when more general cost matrix is given, where not only misclassifications but correct classifications also incur penalties.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.5
• /
• pp.545-551
• /
• 2005

The main objective of fraud detection is to minimize costs or losses that are incurred due to fraudulent transactions. Because of the problem's nature such as highly skewed, overlapping class distribution and non-uniform misclassification costs, it is, however, practically difficult to generate a classifier that is near-optimal in terms of classification costs at a desired operating range of rejection rates. This paper defines a performance measure that reflects classifier's costs at a specific operating range and offers a cost-sensitive learning approach that enables us to train classifiers suitable for real-world credit card fraud detection by directly optimizing the performance measure with evolutionary programming. The experimental results demonstrate that the proposed approach provides an effective way of training cost-sensitive classifiers for successful fraud detection, compared to other training methods.

• Nuclear Engineering and Technology
• /
• v.52 no.7
• /
• pp.1429-1435
• /
• 2020

Applying an accurate parametric prediction model to identify abnormal or false pressurizer water levels (PWLs) is critical to the safe operation of marine pressurized water reactors (PWRs). Recently, deep-learning-based models have proved to be a powerful feature extractor to perform high-accuracy prediction. However, the effectiveness of models still suffers from two issues in PWL prediction: the correlations shifting over time between PWL and other feature parameters, and the example imbalance between fluctuation examples (minority) and stable examples (majority). To address these problems, we propose a cost-sensitive mechanism to facilitate the model to learn the feature representation of later examples and fluctuation examples. By weighting the standard mean square error loss with a cost-sensitive factor, we develop a Cost-Sensitive Long Short-Term Memory (CSLSTM) model to predict the PWL of PWRs. The overall performance of the CSLSTM is assessed by a variety of evaluation metrics with the experimental data collected from a marine PWR simulator. The comparisons with the Long Short-Term Memory (LSTM) model and the Support Vector Regression (SVR) model demonstrate the effectiveness of the CSLSTM.

• Journal of the Korea Institute of Building Construction
• /
• v.12 no.6
• /
• pp.624-635
• /
• 2012

Construction projects, including housing, are carried out over long periods of time. According to changes to the construction period, the cost of input materials and wages also changes. Therefore appropriate management is important in order to minimize cost risks caused by fluctuations in prices. In Korea, housing units are usually sold in lots prior to construction completion. Therefore, careful management of input elements such as materials and equipment that are sensitive to price fluctuations is very important. This study deals with how the price fluctuation of materials, labor, and equipment influences the change of housing cost and seeks a way for cost management through identifying key resources sensitive to price fluctuation. As a result, a change to the housing cost index multiplies depending on cost changes of materials and labor together. Labor costs are a major factor on the housing cost index. In addition, certain types of materials and labor input to housing construction greatly influence price fluctuations. Thus, it is found that managing those main cost factors is the key for effective cost management.

• Journal of Communications and Networks
• /
• v.18 no.5
• /
• pp.806-817
• /
• 2016

Opportunistic routing (OR) has been proposed as a viable approach to improve the performance of wireless multihop networks with lossy links. However, the exponential growth of the bandwidth-sensitive mobile traffic (e.g., mobile video streaming and online gaming) poses a great challenge to the performance of OR in term of bandwidth guarantee. To solve this problem, a novel mechanism is proposed to opportunistically forwarding data packets and provide bandwidth guarantee for the bandwidth-sensitive traffic. The proposal exploits the broadcast characteristic of wireless transmission and reduces the negative effect of wireless lossy links. First, the expected available bandwidth (EAB) and the expected transmission cost (ETC) under OR are estimated based on the local available bandwidth, link delivery probability, forwarding candidates, and prioritization policy. Then, the policies for determining and prioritizing the forwarding candidates is devised by considering the bandwidth and transmission cost. Finally, bandwidth-aware routing algorithm is proposed to opportunistically delivery data packets; meanwhile, admission control is applied to admit or reject traffic flows for bandwidth guarantee. Extensive simulation results show that our proposal consistently outperforms other existing opportunistic routing schemes in providing performance guarantee.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.12 no.1
• /
• pp.6-14
• /
• 2012

All types of part-of-speech (POS) tagging errors have been equally treated by existing taggers. However, the errors are not equally important, since some errors affect the performance of subsequent natural language processing seriously while others do not. This paper aims to minimize these serious errors while retaining the overall performance of POS tagging. Two gradient loss functions are proposed to reflect the different types of errors. They are designed to assign a larger cost for serious errors and a smaller cost for minor errors. Through a series of experiments, it is shown that the classifier trained with the proposed loss functions not only reduces serious errors but also achieves slightly higher accuracy than ordinary classifiers.

• Asia-Pacific Journal of Business
• /
• v.9 no.4
• /
• pp.67-84
• /
• 2018

This study is about comparing coal thermal plant to LNG combined power plant in respect of environmental and economic cost analysis. In addition sensitive analysis of power cost and discount rate is conducted to compare the result of change in endogenous and exogenous variable. For environmental assessment, when they generate 10,669GWh yearly, coal thermal power plant emits sulfur oxides 959ton, nitrogen oxide 690ton, particulate matter 168ton and LNG combined power plant emits only nitrogen oxide 886ton respectively every year. Regarding economic cost analysis on both power plants during persisting period 30 years, coal thermal power plant is more cost effective 4,751 billion won than LNG combined taking in account the initial, operational, energy and environmental cost at 10,669GWh yearly in spite of only LNG combined power plant's energy cost higher than coal thermal. In case of sensitive analysis of power cost and discount rate, as 1% rise or drop in power cost, the total cost of coal thermal power plant increases or decreases 81 billion won and LNG combined 157 billion won up or down respectively. When discount rate 1% higher, the cost of coal thermal and LNG combined power plant decrease 498 billion won and 539 billion won for each. When discount rate 1% lower, the cost of both power plant increase 539 billion won and 837 billion won. With comparing each result of change in power cost and discount rate, as discount rate is weigher than power cost, which means most influential variable of power plan is discount rate one of exogenous variables not endogenous.

• The Journal of Bigdata
• /
• v.6 no.2
• /
• pp.161-168
• /
• 2021

In this study, we propose a cardiovascular disease prediction model using machine learning. First, a multidimensional analysis of various differences between the two groups is performed and the results are visualized. In particular, we propose a predictive model using cost-sensitive learning that can improve the sensitivity for cases where there is a high class imbalance between the normal and patient groups, such as diseases. In this study, a predictive model is developed using CART and XGBoost, which are representative machine learning technologies, and prediction and performance are compared for cardiovascular disease patient data. According to the study results, CART showed higher accuracy and specificity than XGBoost, and the accuracy was about 70% to 74%.