서포트 벡터 머신을 이용한 탄산염암 저류층에서의 암상 예측

서광원,임종세 한국자원공학회 2010 한국자원공학회지 Vol.47 No.2

저류층의 특성변수인 공극률과 유체투과도는 주로 암상의 변화에 의해 결정되며, 물리검층자료로부터 암상을 예측하기 위한 방법으로 경험적인 방법과 추계학적인 방법이 있다. 이 연구에서는 자료 간의 최적 분류 초평면을 결정함으로써 구조적인 위험을 최소화하는 패턴인식 기법인 서포트 벡터 머신을 이용하여 코어 분석 자료와 물리검층자료를 통합적으로 해석함으로써 코어가 취득되지 않은 구간에서의 암상을 예측하였다. 입력 자료로 미국 텍사스 주의 Permian Basin에 위치한 탄산염암 저류층인 Salt Creek Field Unit의 물리검층자료와 코어의 암상 자료를 사용하였으며, 암상 예측 결과를 선형 판별 분석에 의한 암상 예측 결과와 비교하였다. 학습 자료에 대한 교차검증 시 서포트 벡터 머신의 정확도는 평균 98.7%로서 불균질성이 심한 탄산염암 저류층에서도 특징변수에 대한 분류 능력이 우수함을 확인하였다. 예측 결과 5개의 시추공에서 선형 판별 분석의 암상 예측의 정확도는 55.3%이고 서포트 벡터 머신에 의한 암상 예측의 정확도는 평균 67.6%로서 서포트 벡터 머신에 의한 예측이 더 우수함을 확인하였다. 또한 유사한 물리검층 특성을 갖는 암상을 하나의 군집으로 가정하였을 경우에 서포트 벡터 머신이 82.3%의 높은 예측 정확도를 나타내었다. Petrophysical properties such as porosity and permeability are mainly determined by lithofacies variation. There are empirical and stochastic methods to identify lithofacies from well logs. In this study, lithofacies in un-cored ranges are estimated by support vector machine as integrating well log data and core data. Support vector machine is a pattern recognition method minimizing structural risk by optimal separating hyperplane. Well log data and core lithofacies data in carbonate reservoir, Salt Creek Field Unit, Permian basin, Texas, U.S.A., are used as input data. The results of lithofacies estimation by support vector machine are compared with lithofacies estimated by linear discriminant analysis. Average accuracy of cross validation is 98.7% in support vector machine. It is evidence that support vector machine is excellent for classifying characteristic variables in heterogeneous carbonate reservoir. Average accuracy of lithofacies estimation is 67.6% by support vector machine and 55.3% by linear discriminant analysis, respectively. Therefore, it is confirmed that support vector machine is more accurate than linear discriminant analysis. Average accuracy of lithofacies estimation by support vector machine is 82.3% when several lithofacies that have similar characteristics on well log data are assumed as a group.

Multiclass Least Squares Twin Support Vector Machine for Pattern Classification

Divya Tomar,Sonali Agarwal 보안공학연구지원센터 2015 International Journal of Database Theory and Appli Vol.8 No.6

This paper proposes a Multiclass Least Squares Twin Support Vector Machine (MLSTSVM) classifier for multi-class classification problems. The formulation of MLSTSVM is obtained by extending the formulation of recently proposed binary Least Squares Twin Support Vector Machine (LSTSVM) classifier. For M-class classification problem, the proposed classifier seeks M-non parallel hyper-planes, one for each class, by solving M-linear equations. A regularization term is also added to improve the generalization ability. MLSTSVM works well for both linear and non-linear type of datasets. It is relatively simple and fast algorithm as compared to the other existing approaches. The performance of proposed approach has been evaluated on twelve benchmark datasets. The experimental result demonstrates the validity of proposed MLSTSVM classifier as compared to the typical multi-classifiers based on ‘Support Vector Machine’ and ‘Twin Support Vector Machine’. Statistical analysis of the proposed classifier with existing classifiers is also performed by using Friedman’s Test statistic and Nemenyi post hoc techniques.

입자군집 최적화를 이용한 SVM 기반 다항식 뉴럴 네트워크 분류기 설계

노석범(Seok-Beom Roh),오성권(Sung-Kwun Oh) 대한전기학회 2018 전기학회논문지 Vol.67 No.8

In this study, the design methodology as well as network architecture of Support Vector Machine based Polynomial Neural Network, which is a kind of the dynamically generated neural networks, is introduced. The Support Vector Machine based polynomial neural networks is given as a novel network architecture redesigned with the aid of polynomial neural networks and Support Vector Machine. The generic polynomial neural networks, whose nodes are made of polynomials, are dynamically generated in each layer-wise. The individual nodes of the support vector machine based polynomial neural networks is constructed as a support vector machine, and the nodes as well as layers of the support vector machine based polynomial neural networks are dynamically generated as like the generation process of the generic polynomial neural networks. Support vector machine is well known as a sort of robust pattern classifiers. In addition, in order to enhance the structural flexibility as well as the classification performance of the proposed classifier, multi-objective particle swarm optimization is used. In other words, the optimization algorithm leads to sequentially successive generation of each layer of support vector based polynomial neural networks. The bench mark data sets are used to demonstrate the pattern classification performance of the proposed classifiers through the comparison of the generalization ability of the proposed classifier with some already studied classifiers.

Linear One-Class Support Tensor Machine

Yanyan Chen,Ping Zhong 보안공학연구지원센터 2016 International Journal of Signal Processing, Image Vol.9 No.9

One-class support vector machine is an important and efficient classifier which is used in the situation that only one class of data is available, and the other is too expensive or difficult to collect. It uses vector as input data, and trains a linear or nonlinear decision function in vector space. However, there is reason to consider data as tensor. Tensor representation can make use of the structural information present in the data, which cannot be handled by the traditional vector based classifier. The significant benefit of using tensor as input is the reduction of the number of decision parameters, which can avoid the overfitting problems and especially suitable for small sample and large dimension cases. In this paper we have proposed a tensor based one-class classification algorithm named linear one-class support tensor machine. It aims to find a hyperplane in tensor space with maximal margin from the origin that contains almost all the data of the target class. We demonstrate the performance of the new tensor based classifier on several publicly available datasets in comparison with the standard linear one-class support vector machine. The experimental results indicate the validity and advantage of our tensor based classifier.

퍼지 원 클래스 서포트 백터 머신

김기주 ( Kijoo Kim ),최영식 ( Youngsik Choi ) 한국인터넷정보학회 2005 인터넷정보학회논문지 Vol.6 No.3

OC-SVM(One Class Support Vector Machine)은 주어진 전체 데이터의 분포를 측정하는 대신에, 데이터 분포의 서포트(support)를 측정하는 기술로서 주어진 데이터를 가장 잘 설명할 수 있는 최적의 서포트 벡터(support vector)를 구하는 기술이다. OC-SVM은 데이터 분포의 표현에 아주 뛰어난 접근 방법이지만, 사람의 주관적인 중요도를 반영하는 것은 힘들다. 본 논문에서는 각 데이터에 퍼지 맴버쉽(fuzzy membership)을 적용하여 기존의 OC-SVM에 사용자의 주관적인 중요도를 표현할 수 있는 FOC-SVM(Fuzzy One Class Support Vector Machine)을 유도 하였다. FOC-SVM은 데이터들을 동등하게 다루는 것이 아니라, 데이터 객체의 중요도에 따라 데이터를 다룬다. 즉, 덜 중요한 데이터의 특징 벡터는 OC-SVM의 처리과정에 덜 기여하도록 하기 위하여, 객체의 중요도에 따라 특징 벡터의 크기를 조정하였다. 이를 증명하기 위하여 가상의 데이터를 가지고 실험을 하였고, 실험 결과는 예측된 결과를 보여 주었다. OC-SVM(One Class Support Vector Machine) avoids solving a full density estimation problem, and instead focuses on a simpler task, estimating quantiles of a data distribution, i.e. its support. OC-SVM seeks to estimate regions where most of data resides and represents the regions as a function of the support vectors. Although OC-SVM is powerful method for data description, it is difficult to incorporate human subjective importance into its estimation process. In order to integrate the importance of each point into the OC-SVM process, we propose a fuzzy version of OC-SVM. In FOC-SVM (Fuzzy One-Class Support Vector Machine), we do not equally treat data points and instead weight data points according to the importance measure of the corresponding objects. That is, we scale the kernel feature vector according to the importance measure of the object so that a kernel feature vector of a less important object should contribute less to the detection process of OC-SVM. We demonstrate the performance of our algorithm on several synthesized data sets. Experimental results showed the promising results.

희박한 최소 절대 편차 지지벡터기계

정강모(Kang-Mo Jung) 한국자료분석학회 2023 Journal of the Korean Data Analysis Society Vol.25 No.5

고전적인 지지기계벡터가 어떤 부등식 제약 조건에서 최적화 문제의 해를 구하는 것에 비해 최소 제곱 지지기계벡터는 이 부등식 제약 조건을 등식 제약 조건으로 변환하여 문제의 해를 구한다. 따라서 최소 제곱 지지기계벡터는 행렬을 이용하여 정확 해를 구할 수 있어 회귀와 분류문제의 많은 분야에서 탁월한 성과를 이뤘다. 그러나 최소 제곱 지지기계벡터에서 구한 해는 이상치에 민감하고, 고전적인 지지기계벡터의 장점인 희박한 지지벡터를 제공하지 못한다는 단점이 있다. 이를 해결하기 위해 본 논문에서는 최소 절댓값 손실함수를 이용함으로써 이상치에 강건한 최소 절대 편차 지지기계벡터의 해를 구한다. 또한, 지지벡터의 희박성을 위해 재귀적 축소 최소 제곱 지지기계벡터를 이용하는 방법을 제시하고자 한다. 최소 절댓값 손실함수의 최적화 문제를 해결하기 위해 분리-브레그만 반복 방법을 사용하여 정확한 해를 구하였다. 본 논문에서 제시한 방법은 기존의 최소 제곱 지지기계벡터가 가지는 단점을 극복하는 효율적인 방법으로 간단한 수치 자료와 벤치마크 자료의 분석 결과가 해의 강건성과 희박성 측면에서 기존 결과와 비교할 만한 수준을 보였다. The support vector machine solves a quadratic programming problem with linear inequality and equality constraints. However, it is not trivial to solve the quadratic problem. The least squares support vector machine(LS-SVM) solves a linear system by equality constraints instead of inequality constraints. LS-SVM is a popular method in regression and classification problems, because it effectively solves simple linear systems. There are two issues with the LS-SVM solution : the lack of robustness to outliers and the absence of sparseness. In this paper, we propose a sparse and robust support vector machine for regression problems using the least absolute deviation support vector machine (LAD-SVM) and recursive reduced LS-SVM (RR-LS-SVM). The split-Bregman iteration gives the exact solution for the LAD-SVM problem, while RR-LS-SVM gives a sparse solution with a much smaller number of all support vectors. Numerical experiments with simulation and benchmark data demonstrate that the proposed algorithm can achieve comparable performance to other methods in terms of robustness and sparseness.

A Kernel-based Matrixzed One-Class Support Vector Machine

Yanyan Chen,June Yuan,Zhengkun Hu 보안공학연구지원센터 2016 International Journal of Hybrid Information Techno Vol.9 No.11

One-class support vector machine (OCSVM) is an important and efficient classifier used when only one class of data is available while others are too expensive or difficult to collect. It uses vector as input data, and trains a linear or nonlinear decision function in vector space. However, the traditional vector-based classifiers may fail when input is matrix. Therefore, it makes sense to study matrixzed classifiers which can make use of the structural information presented in the data. In this paper we propose a matrix-based one-class classification algorithm named Kernel-based Matrixzed One-class Support Vector Machine (KMatOCSVM). It aims to convert the OCSVM to suit for matrix representation data and to deal with nonlinear one-class classification problems. The efficiency and validity of the proposed method is illustrated by four real-world matrix-based human face datasets.

An Application of Support Vector Machines to Personal Credit Scoring : Focusing on Financial Institutions in China

Xuan-Ze Ding,Young-Chan Lee 대한산업경영학회 2018 산업융합연구 Vol.16 No.4

개인신용평가는 은행이 대출을 승인할 때 수익성 있는 의사결정을 적절히 유도할 수 있는 효과적인 도구이다. 최근 많은 분류 알고리즘 및 모델이 개인신용평가에 사용되고 있다. 개인신용평가 기법은 대체로 통계적 방법과 비 통계 적 방법으로 구분된다. 통계적 방법에는 선형회귀분석, 판별분석, 로지스틱 회귀분석, 의사결정나무 등이 포함된다. 비 통 계적 방법에는 선형계획법, 신경망, 유전자 알고리즘 및 Support Vector Machines 등이 포함된다. 그러나 신용평가모형 개발을 위해 어떠한 방법이 최선인지에 관해서는 일관된 결론을 내리기는 어렵다. 본 논문에서는 중국 금융기관의 개인 신용 데이터를 사용하여 가장 대표적인 신용평가 기법인 로지스틱 회귀분석, 신경망 그리고 Support Vector Machines의 성능을 비교하고자 한다. 구체적으로, 세 가지 모형을 각각 구축하여 고객을 분류하고 분석 결과를 비교하였다. 분석결과 에 따르면, Support Vector Machines이 로지스틱 회귀분석과 신경망보다 더 나은 성능을 가지는 것으로 나타났다. Personal credit scoring is an effective tool for banks to properly guide decision profitably on granting loans. Recently, many classification algorithms and models are used in personal credit scoring. Personal credit scoring technology is usually divided into statistical method and non-statistical method. Statistical method includes linear regression, discriminate analysis, logistic regression, and decision tree, etc. Non-statistical method includes linear programming, neural network, genetic algorithm and support vector machine, etc. But for the development of the credit scoring model, there is no consistent conclusion to be drawn regarding which method is the best. In this paper, we will compare the performance of the most common scoring techniques such as logistic regression, neural network, and support vector machines using personal credit data of the financial institution in China. Specifically, we build three models respectively, classify the customers and compare analysis results. According to the results, support vector machine has better performance than logistic regression and neural networks.

이미지 데이터에 대한 분류 방법의 비교 연구

박범진(Beom-Jin Park),박창이(Changyi Park) 한국데이터정보과학회 2018 한국데이터정보과학회지 Vol.29 No.1

이미지는 행렬형태로 자연스럽게 표현되므로 기존의 기계 학습 (machine learning) 방법들을 이미지 데이터에 적용하기 위해서는 행렬을 벡터로 변환해야 한다. 최근 지지행렬기계 (support matrix machine)는 데이터 행렬을 벡터로 변환하지 않고 직접 분류하도록 고안되었다. 그러나 문헌상의 연구에서는 지지행렬기계와 지지벡터기계 (support vector machine)의 분류 정확도만을 비교하였다. 본 논문에서는 지지벡터기계의 예측 성능을 k-근방 분류, 지지벡터기계, 그리고 심층 신경망 (deep neural network)과 같은 이미지 데이터에 대한 주요 분류방법들과 비교하고 이러한 방법들의 특징에 대하여 알아보고자 한다. Since images are naturally represented as matrices, we have to reshape matrices into vectors in order to apply traditional methods in machine learning to image data. Recently, support matrix machine (SMM) has been proposed to directly classify data matrices without reshaping those matrices into vectors. However, the classification accuracies of SMM and support vector machine were compared in the literature. In this paper, we compare the predictive performance of SMM with those of major classification methods for image data such as k-nearest neighborhood classifier, support vector machine, and deep neural network and understand the characteristics of those learning methods.

First Principle을 결합한 최소제곱 Support Vector Machine의 예측 능력

김병주,심주용,황창하,김일곤 한국정보과학회 2003 정보과학회논문지 : 소프트웨어 및 응용 Vol.30 No.7

본 논문에서는 최근 뛰어난 예측력으로 각광받는 최소제곱 Support Vector Machine(Least Square Support Vector Machine: LS-SVM)과 First Principle(FP)을 결합한 하이브리드 최소제곱ㆍSupport Vector Machine 모델, HLS-SVM(Hybrid Least Square-Super Vector Machine)을 제안한다. 제안한 모델인 하이브리드 최소제곱 Support Vector Machine을 기존의 방법인 하이브리드 신경망(Hybrid Neural Network:HNN), 비선형 칼만필터와 하이브리드 신경망을 결합한 HNN-EKF (Hybrid Neural Network with Extended Kalman Filter) 모델과 비교해 보았다. HLS-SVM 모델은 학습 및 validation 과정에서는 HNN-EKF와 근사한 성능을 보였고, HNN 보다는 우수한 결과를 보였고, 일반화 성능에서는 HNN-EKF에 비해 3배, HNN보다 100배정도 우수한 결과를 보였다. A hybrid least square Support Vector Machine combined with First Principle(FP) knowledge is proposed. We compare hybrid least square Support Vector Machine(HLS-SVM) with early proposed models such as Hybrid Neural Network(HNN) and HNN with Extended Kalman Filter(HNN-EKF). In the training and validation stage HLS-SVM shows similar performance with HNN-EKF but better than HNN, whereas, in the testing stage, it shows three times better than HNN-EKF, hundred times better than HNN model.