• Title/Summary/Keyword: 선형부정

Search Result 122, Processing Time 0.026 seconds

Fraud detection support vector machines with a functional predictor: application to defective wafer detection problem (불량 웨이퍼 탐지를 위한 함수형 부정 탐지 지지 벡터기계)

  • Park, Minhyoung;Shin, Seung Jun
    • The Korean Journal of Applied Statistics
    • /
    • v.35 no.5
    • /
    • pp.593-601
    • /
    • 2022
  • We call "fruad" the cases that are not frequently occurring but cause significant losses. Fraud detection is commonly encountered in various applications, including wafer production in the semiconductor industry. It is not trivial to directly extend the standard binary classification methods to the fraud detection context because the misclassification cost is much higher than the normal class. In this article, we propose the functional fraud detection support vector machine (F2DSVM) that extends the fraud detection support vector machine (FDSVM) to handle functional covariates. The proposed method seeks a classifier for a function predictor that achieves optimal performance while achieving the desired sensitivity level. F2DSVM, like the conventional SVM, has piece-wise linear solution paths, allowing us to develop an efficient algorithm to recover entire solution paths, resulting in significantly improved computational efficiency. Finally, we apply the proposed F2DSVM to the defective wafer detection problem and assess its potential applicability.

Fuzzy Disturbance Observer based Multiple Sliding Surface Control of Nonlinear Systems with Mismatched Disturbance (부정합조건 외란을 갖는 비선형 시스템의 퍼지 외란 관측기 기반 다중 슬라이딩 평면 제어)

  • Lee, Sang-Yun;Seo, Hyungkeun;Hyun, Chang-Ho;Park, Mignon
    • Journal of the Korean Institute of Intelligent Systems
    • /
    • v.24 no.4
    • /
    • pp.385-391
    • /
    • 2014
  • This paper proposes fuzzy disturbance observer based multiple sliding surface control scheme for nonlinear systems with mismatched disturbance. In order to stabilize nonlinear systems with mismatched disturbance, a controller based on multiple sliding surface control scheme is designed. In addition, a fuzzy disturbance observer is used to estimate the disturbance. Using the fuzzy disturbance observer, "explosion of terms" problem and chattering problem were solved. The stability of the proposed control scheme is analyzed by Lyapunov stability theory. For the verification, we apply the proposed method to numerical examples and compare its result with that of the applied nonlinear disturbance observer based sliding mode control.

Hydraulic Flood Routing using Linear Reservoir Model (선형저수지모형을 적용한 수리학적 홍수추적)

  • Jeon, Min-Woo;Cho, Young-Soo
    • Journal of Korea Water Resources Association
    • /
    • v.35 no.6
    • /
    • pp.787-796
    • /
    • 2002
  • Hydraulic flood routing was performed for unsteady flow in a natural river using Preissmann scheme. A Log-Pearson Type-Ⅲ hydrograph is chosen arbitrarily as the upstream boundary condition and lateral inflow hydrographs for sensitivity analysis. For the application with an actual river system, upstream and lateral inflow hydrographs were estimated by the linear reservoir model and the Manning's equation was used as the downstream boundary condition. The unsteady flow model using the linear reservoir model as the inflow hydrographs was applied to Bochung stream basin and gives good results, and is approved to be used for the runoff prediction. As results of the sensitivity analysis, the proposed model may help to estimate the roughness coefficients when using the unsteady flow model with lateral inflow combined with the linear reservoir model.

A bicoherence analysis of EEG during Yes/No decision task (긍/부정 문답 과제 수행시 뇌파의 바이코히어런스 분석)

  • 남승훈;류창수;임태규;송윤선;유창용
    • Proceedings of the Korean Society for Emotion and Sensibility Conference
    • /
    • 2003.05a
    • /
    • pp.115-119
    • /
    • 2003
  • 본 연구는 인간에 있어서 가장 간단한 의사라고 여겨지는 긍정과 부정 응답에 대해 나타나는 뇌파의 변화를 잘 반영하는 특징을 찾아내고자 하기 위한 것이다. 고차 통계적 방법(high order statical analysis)인 바이스펙트럼(bispectrum)은 뇌파의 다른 부위와 다른 주파수 사이의 비선형위상커플링(non-linear phase coupling)을 잘 반영하므로, 이를 이용하여 긍정이나 부정을 선택할 때 나타나는 뇌파를 분석하였다. 분석결과, 반응 전 1.25초∼0.5초 에 유의미한 차이를 보였다. 긍정과 부정 응답에 대한 뇌파의 주파수와 부위를 찾아 신경회로망의 입력으로 사용하여 긍정과 부정 응답에 대해 분별하였다. 2번의 뇌파실험에서 각각 실험 데이터에 대해서는 긍/부정 차이가 존재하지만 공통적인 특징이 나타나지는 않았다.

  • PDF

A Study on the Nonlinear Buckling Behavior of Thin-Walled Sections (박판단면의 비선형 좌굴거동에 관한 해석적연구)

  • Jin, Chang Sun;Kwon, Young Bong
    • Journal of Korean Society of Steel Construction
    • /
    • v.10 no.3 s.36
    • /
    • pp.407-421
    • /
    • 1998
  • The purpose of this paper is to provide and verify an analytical method, based on the spline finite strip method, which can be used to investigate the buckling mode and stress of thin-walled steel sections. Geometric imperfection and initial stress of plates and plate assemblies, which are resulted from various preloadings and may cause prebuckling deformations before buckling, are included in the analysis. Material nonlinearity and residual stress are also considered. It can be applied to sections with simple or non-simple boundary conditions and arbitrary loading. The method has been applied to investigate the buckling behavior of plates and plate assemblies which are subjected to compression with initial imperfections and residual stresses.

  • PDF

The syntax of Linear logic (선형논리의 통사론)

  • Cheong, Kye-Seop
    • Journal for History of Mathematics
    • /
    • v.25 no.3
    • /
    • pp.29-39
    • /
    • 2012
  • As a product of modern proof theory, linear logic is a new form of logic developed for the purpose of enhancing programming language by Professor Jean-Yves Girard of Marseille University (France) in 1987 by supplementing intuitionist logic in a sophisticated manner. Thus, linear logic' s connectives can be explained using information processing terms such as sequentiality and parallel computation. For instance, A${\otimes}$B shows two processes, A and B, carried out one after another. A&B is linked to an internal indeterminate, allowing an observer to select either A or B. A${\oplus}$B is an external indeterminate, and as such, an observer knows that either A or B holds true, but does not know which process will be true. A ${\wp}$ B signifies parallel computation of process A and process B; linear negative exhibits synchronization, that is, in order for the process A to be carried out, both A and $A^{\bot}$ have to be accomplished simultaneously. Since the field of linear logic is not very active in Korea at present, this paper deals only with syntax aspect of linear logic in order to arouse interest in the subject, leaving semantics and proof nets for future studies.

A New Robust Continuos VSCS by Saturation Function for Uncertain Nonlinear Plants (불확실 비선형 플랜트를 위한 포화 함수에 의한 새로운 강인한 연속 가변구조제어시스템)

  • Lee, Jung-Hoon
    • Journal of the Institute of Electronics Engineers of Korea SC
    • /
    • v.48 no.3
    • /
    • pp.30-39
    • /
    • 2011
  • In this note, a systematic design of a new robust nonlinear continuous variable structure control system(VSCS) based on the modified state dependent nonlinear form is presented for the control of uncertain affine nonlinear systems with mismatched uncertainties and matched disturbance. After an affine uncertain nonlinear system is represented in the form of state dependent nonlinear system, a systematic design of a new robust nonlinear VSCS is presented. The uncertainty of the nonlinear system function is separated into the tow parts, i.e., state dependent term and state independent term for extension of target plants. To be linear in the closed loop resultant dynamics and in order to easily satisfy the existence condition of the sliding mode, the transformed linear sliding surface is applied. A corresponding control input is proposed to satisfy the closed loop exponential stability and the existence condition of the sliding mode on the linear transformed sliding surface, which will be investigated in Theorem 1. For practical application, the discontinuity of the control input as the inherent property of the VSS is improved dramatically. Through a design example and simulation studies, the usefulness of the proposed controller is verified.

Digital Calibration Technique for Cyclic ADC based on Digital-Domain Averaging of A/D Transfer Functions (아날로그-디지털 전달함수 평균화기법 기반의 Cyclic ADC의 디지털 보정 기법)

  • Um, Ji-Yong
    • Journal of the Institute of Electronics and Information Engineers
    • /
    • v.54 no.6
    • /
    • pp.30-39
    • /
    • 2017
  • A digital calibration technique based on digital-domain averaging for cyclic ADC is proposed. The proposed calibration compensates for nonlinearity of ADC due to capacitance mismatch of capacitors in 1.5-bit/stage MDAC. A 1.5-bit/stage MDAC with non-matched capacitors has symmetric residue plots with respect to the ideal residue plot. This intrinsic characteristic of residue plot of MDAC is reflected as symmetric A/D transfer functions. A corrected A/D transfer function can be acquired by averaging two transfer functions with non-linearity, which are symmetric with respect to the ideal analog-digital transfer function. In order to implement the aforementioned averaging operation of analog-digital transfer functions, a 12-bit cyclic ADC of this work defines two operational modes of 1.5-bit/stage MDAC. By operating MDAC as the first operational mode, the cyclic ADC acquires 12.5-bits output code with nonlinearity. For the same sampled input analog voltage, the cyclic ADC acquires another 12.5-bits output code with nonlinearity by operating MDAC as the second operational mode. Since analog-digital transfer functions from each of operational mode of 1.5-bits/stage MDAC are symmetric with respect to the ideal analog-digital transfer function, a corrected 12-bits output code can be acquired by averaging two non-ideal 12.5-bits codes. The proposed digital calibration and 12-bit cyclic ADC are implemented by using a $0.18-{\mu}m$ CMOS process in the form of full custom. The measured SNDR(ENOB) and SFDR are 65.3dB (10.6bits) and 71.7dB, respectively. INL and DNL are measured to be -0.30/-0.33LSB and -0.63/+0.56LSB, respectively.

熔接構造物 의 不整量 에 대하여

  • 등전양
    • Journal of Welding and Joining
    • /
    • v.1 no.1
    • /
    • pp.21-29
    • /
    • 1983
  • 용접구조물에는 필연적으로 용접에 의해서 생기는 부정량이 존재한다. 용접에 의해서 건조되는 선체구조에 있어서도 misalignment를 비롯하여 과대한 gap, 팬널(panel)의 초기변형, 용접에 의한 수축과 같은 부정량이 존재하여 선속 불록의 정도와 선체 강도에 영향을 주게 된다. 그래서 일본에서는 건조되는 동선의 품질을 고도로 유지하면서 공수를 될 수 있는 한 줄이기 위하여 이론과 실속이 뒷받침된 신속도가 높은 품질표준을 설정하는 것이 필요하다고 판단하여 1964년에 '일본동선공작법표준, japan shipbuilding quality standard(JSQS)'를 제정하였다. 그 내용에는 소재, 현도, 내용 가공부재, 조립 블록, 선형확보, 다듬질, 변형량등 주로 공정순에 따라 각각 구 분을 정한 항목마다 표준범위와 허용 한계가 표시되어 있다. 지금부터 용접에 의한 부정량의 대표예로써, 용접에 의한 변형과 잔유응력, 이음에서의 gap과 misalignment에 대하여 말하고 아울러 부정량과 JSQS와의 관계에 대해서도 간단히 말하겠다.

  • PDF

Traitor Traceability of Colluded Multimedia Fingerprinting code Using Hamming Distance on XOR Collusion Attack (XOR 공모공격에서 해밍거리를 이용한 공모된 멀티미디어 핑거프린팅 코드의 부정자 추적)

  • Chung, Il Yong;Rhee, Kang Hyeon
    • Journal of the Institute of Electronics and Information Engineers
    • /
    • v.50 no.7
    • /
    • pp.175-180
    • /
    • 2013
  • For the traitor tracing of multimedia content, this paper presents the classification algorithm of XOR collusion attack types using hamming distance, which applies to the colluded fingerprinting codes. The conventional traitor decision hinges on the colluded fingerprinting code used by a correlation coefficient, but the proposed scheme uses hamming distance. While XOR collusion attack employing a correlation coefficient is impossible to trace the traitors about 50% colluders due to a serious XOR linear problem, our method improves the performance of traceability to trace at least 1 traitor using hamming distance, and thus, the functional behavior of the proposed traitor traceability is coincided with Probability Scheme.