• Journal of Korean Society for Quality Management
• /
• v.17 no.1
• /
• pp.35-47
• /
• 1989

This paper is concerned with a development and evaluation of heuristic algorithms for the n-job, M-stage flowshop with sequence dependent setup times. Three heuristic algorithms, CAIDAN, DANNEN and PETROV, are proposed. The makespan is taken as a performance measure for the algorithms. The experiment for each algorithm is designed for a $4{\times}3{\times}3$ factorial design with 360 observations. The experimental factors are PS (ratio of processing times to setup times), M (number of machines), and N (number of jobs). The makespan of the proposed heuristic algorithms is compared with the optimal makespan obtained by the complete enumeration method. The result of comparision of performance measure is called a relative error. The mean relative errors of CAIDAN, DANNEN and PETROV algorithms are 4.488%. 6.712% and 7.282%, respectively. The computational results are analysed using SPSS. The experimental results show that the three factors are statistically signiticant at 5% level.

• Proceedings of the KSRS Conference
• /
• 1998.09a
• /
• pp.213-217
• /
• 1998

This study overviews existing evaluation methods of classification accuracy using confusion matrix proposed by Cohen in 1960's, and proposes ISDd(Index of Spatial Distribution by distance) and ISDs(Index of Spatial Distribution by scatteredness) for the evaluation of spatial distribution of satellite image classification errors, which has not been tried yet. Index of spatial distribution offers the basis of decision on adoption/rejection of classification results at sub-image level by evaluation of distribution, such as status of local aggregation of misclassified pixels. So, users can understand the spatial distribution of misclassified pixels and, can have the basis of judgement of suitability and reliability of classification results.

• The Korean Journal of Applied Statistics
• /
• v.15 no.1
• /
• pp.187-199
• /
• 2002

When the sample size is small, exact tests are often employed because the asymptotic distribution of the test statistic is in doubt. The advantage of exact tests is that it is guaranteed to bound the type I error probability to the nominal level. In this paper we review the methods of constructing exact tests, the algorithm and commercial software. We also examine the difference between exact p-values obtained from exact tests and true p-values obtained from the true underlying distribution.

• Proceedings of the Korean Statistical Society Conference
• /
• 2005.05a
• /
• pp.7-12
• /
• 2005

In this paper, we review the variable-selection properties of LASSO and SCAD in penalized regression. To improve the weakness of SCAD for high noise level, we propose a new penalty function called MSCAD which relaxes the unbiasedness condition of SCAD. In order to compare MSCAD with LASSO and SCAD, comparative studies are performed on simulated datasets and also on a real dataset. The performances of penalized regression methods are compared in terms of relative model error and the estimates of coefficients. The results of experiments show that the performance of MSCAD is between those of LASSO and SCAD as expected.

• Journal of KSNVE
• /
• v.2 no.2
• /
• pp.99-106
• /
• 1992

The complex Young's modulus of a viscoelastic material can be obtained as a function of frequency from the measurements of relative motion between the two ends of a bar-type specimen. Non-resonance method is usually used to obtain the complex Young's modulus over wide range of frequency including resonance points, while in resonance method information at resonance frequencies only is used. However, the complex Young's modulus obtained by the non-resonance method is often unreliable in the anti-resonance frequency regions because of the measurement noise problems. In this study, the effects of the random measurement errors on estimating the complex Young's modulus are studied in the aspect of sensitivity, and how to obtain the reliable frequency region for a given measurement error level is shown. The usable frequency regions in determining the complex Young's modulus are represented by a non-dimensional parameter formed with the wave length and specimen length.

• Journal of Engineering Education Research
• /
• v.16 no.5
• /
• pp.18-23
• /
• 2013

A curriculum is a measure of how high level of knowledge is educated to students affiliated to university institutions, and is used as an indicator evaluating usefulness of the university institutions. For this reason, Designing and assessing curricula is a critical to high education institutions. However, in the case of development and quality assessment of a curriculum with a traditional way, finding a right curriculum which a designer intends is a time consuming and error-prone process. Therefore, in order to improve these problems, we propose the curriculum design engine using SES (System Entity Structure) / DEVS (Discrete Event System Specification) Framework in this paper. The SES describes all possible combination of curricula for students. The DEVS framework provides a simulation environment for models created from the SES by the students. The proposed engine will show appropriate curricula for students after a simulator appropriately filters students' requirement, according to conditions.

• Transactions of the Society of Information Storage Systems
• /
• v.1 no.2
• /
• pp.137-142
• /
• 2005

In this paper, we propose a new servo system design strategy to reduce the position error signal(PES) and track mis-registration(TMR) in magnetic disk drive systems. The proposed method provides a systematic design procedure based on the plant model and an optimal solution via an optimization with a 'Robust Random Neighborhood Search(RRNS)' algorithm. In addition, it guarantees the minimum PES level as well as stability to parametric uncertainties. Furthermore, the proposed method can be used to estimate the performance at the design stage and thus can reduce the cost and time for the design of the next generation product. The reduction of PES as well as robust stability is demonstrated by simulation and experiments.

• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10d
• /
• pp.622-624
• /
• 2002

생체인식 기술은 급속도로 발전하고 있지만 개개의 생체 정보를 이용한 단일 생체인식 기술은 생체 방식에 따라 각각의 문제점이 노출되고 있는 상황이다. 이에 두 가지 이상의 생체 정보를 결합하여 단일 생체인식 기술의 문제점을 극복하고 보다 좋은 인식률을 확보하기 위해 다중 생체인식 시스템(Multi-Modal Bio-metries System)이라는 복합 시스템이 제안 되었다. 이 논문에서는 생체인식 산업의 특성 및 개인 인증 방법으로 사용중인 단일 생체인식 시스템의 문제점을 알아보고 그 해결방안으로 다중 생체인식 시스템의 확률단계(Probability Level)에서 더 좋은 성능을 보여주기 위해 각각의 시스템에 가중치(Weight)를 부여 할 경우, EER(Equal Error Rate)이 단일 생체인식 시스템에 보다 가중치를 부여 했을 때 낮아짐과 동시에 ROC 커브도 (Receiver Operating Characteristic Curve) 좋아짐을 보였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.8 no.6
• /
• pp.2186-2196
• /
• 2014

Spam mail is one of the most general mail dysfunctions, which may cause psychological damage to internet users. As internet usage increases, the amount of spam mail has also gradually increased. Indiscriminate sending, in particular, occurs when spam mail is sent using smart phones or tablets connected to wireless networks. Spam mail consists of approximately 68% of mail traffic; however, it is believed that the true percentage of spam mail is at a much more severe level. In order to analyze and detect spam mail, we introduce a technique based on spam mail characteristics and text mining; in particular, spam mail is detected by extracting the linguistic analysis and language processing. Existing spam mail is analyzed, and hidden spam signatures are extracted using text clustering. Our proposed method utilizes a text mining system to improve the detection and error detection rates for existing spam mail and to respond to new spam mail types.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.8
• /
• pp.3790-3803
• /
• 2018

Content-based image retrieval is an approach used to query images based on their semantics. Semantic based retrieval has its application in all fields including medicine, space, computing etc. Semantically generated binary hash codes can improve content-based image retrieval. These semantic labels / binary hash codes can be generated from unlabeled data using convolutional autoencoders. Proposed approach uses semi-supervised deep hashing with semantic learning and binary code generation by minimizing the objective function. Convolutional autoencoders are basis to extract semantic features due to its property of image generation from low level semantic representations. These representations of images are more effective than simple feature extraction and can preserve better semantic information. Proposed activation and loss functions helped to minimize classification error and produce better hash codes. Most widely used datasets have been used for verification of this approach that outperforms the existing methods.