Transductive support vector machine optimization problem is a NP problem, in the case of larger number of labeled samples, it is often difficult to obtain a global optimal solution, thereby the good generalization abi...Transductive support vector machine optimization problem is a NP problem, in the case of larger number of labeled samples, it is often difficult to obtain a global optimal solution, thereby the good generalization ability of transductive learning has been affected. Previous methods can not give consideration to both running efficiency and classification precision. In this paper, a transductive support vector machine algorithm based on ant colony optimization is proposed to overcome the drawbacks of the previous methods. The proposed algorithm approaches the approximate optimal solution of Transductive support vector machine optimization problem by ant colony optimization algorithm, and the advantage of transductive learning can be fully demonstrated. Experiments on several UCI standard datasets and the newsgroups 20 dataset showed that, with respect to running time and classification precision, the proposed algorithm has obvious advantage over the previous algorithms.展开更多
针对现有多类支持向量域数据描述(multi-class support data description,M-SVDD)存在的不足和局限性,提出一种拒绝式转导推理多类支持向量域数据描述(rejected transductive inference M-SVDD,RTIM-SVDD)方法,并将该方法应用于机械加...针对现有多类支持向量域数据描述(multi-class support data description,M-SVDD)存在的不足和局限性,提出一种拒绝式转导推理多类支持向量域数据描述(rejected transductive inference M-SVDD,RTIM-SVDD)方法,并将该方法应用于机械加工故障诊断当中。首先,RTIM-SVDD通过训练寻求一个尽可能包含所有数据样本的最小超球体作为拒绝检测面,来整体学习样本知识或数据描述,并以一定的拒绝度判别新的测试样本,达到故障检测的目的;其次,应用分别包含各个类别样本的多个超球体,来判别满足一定拒绝度的样本,实现多分类问题。对于模糊样本点归属判别关键问题,本文采用一种新的转导推理规则来进行决策。最后,通过一个仿真实验进行验证,结果证明了RTIM-SVDD的可行性和有效性。展开更多
We advance here a novel methodology for robust intelligent biometric information management with inferences and predictions made using randomness and complexity concepts. Intelligence refers to learning, adap- tation,...We advance here a novel methodology for robust intelligent biometric information management with inferences and predictions made using randomness and complexity concepts. Intelligence refers to learning, adap- tation, and functionality, and robustness refers to the ability to handle incomplete and/or corrupt adversarial information, on one side, and image and or device variability, on the other side. The proposed methodology is model-free and non-parametric. It draws support from discriminative methods using likelihood ratios to link at the conceptual level biometrics and forensics. It further links, at the modeling and implementation level, the Bayesian framework, statistical learning theory (SLT) using transduction and semi-supervised lea- rning, and Information Theory (IY) using mutual information. The key concepts supporting the proposed methodology are a) local estimation to facilitate learning and prediction using both labeled and unlabeled data;b) similarity metrics using regularity of patterns, randomness deficiency, and Kolmogorov complexity (similar to MDL) using strangeness/typicality and ranking p-values;and c) the Cover – Hart theorem on the asymptotical performance of k-nearest neighbors approaching the optimal Bayes error. Several topics on biometric inference and prediction related to 1) multi-level and multi-layer data fusion including quality and multi-modal biometrics;2) score normalization and revision theory;3) face selection and tracking;and 4) identity management, are described here using an integrated approach that includes transduction and boosting for ranking and sequential fusion/aggregation, respectively, on one side, and active learning and change/ outlier/intrusion detection realized using information gain and martingale, respectively, on the other side. The methodology proposed can be mapped to additional types of information beyond biometrics.展开更多
基金This work is sponsored by the National Natural Science Foundation of China (Nos. 61402246, 61402126, 61370083, 61370086, 61303193, and 61572268), a Project of Shandong Province Higher Educational Science and Technology Program (No. J15LN38,J14LN31), Qingdao indigenous innovation program (No. 15-9-1-47-jch), the Project of Shandong Provincial Natural Science Foundation of China (No. ZR2014FL019), the Open Project of Collaborative Innovation Center of Green Tyres & Rubber (No. 2014GTR0020), the National Research Foundation for the Doctoral Program of Higher Education of China (No.20122304110012), the Science and Technology Research Project Foundation of Heilongjiang Province Education Department (No. 12531105), Heilongjiang Province Postdoctoral Research Start Foundation (No. LBH-Q13092), and the National Key Technology R&D Program of the Ministry of Science and Technology under Grant No. 2012BAH81F02.
文摘Transductive support vector machine optimization problem is a NP problem, in the case of larger number of labeled samples, it is often difficult to obtain a global optimal solution, thereby the good generalization ability of transductive learning has been affected. Previous methods can not give consideration to both running efficiency and classification precision. In this paper, a transductive support vector machine algorithm based on ant colony optimization is proposed to overcome the drawbacks of the previous methods. The proposed algorithm approaches the approximate optimal solution of Transductive support vector machine optimization problem by ant colony optimization algorithm, and the advantage of transductive learning can be fully demonstrated. Experiments on several UCI standard datasets and the newsgroups 20 dataset showed that, with respect to running time and classification precision, the proposed algorithm has obvious advantage over the previous algorithms.
文摘针对现有多类支持向量域数据描述(multi-class support data description,M-SVDD)存在的不足和局限性,提出一种拒绝式转导推理多类支持向量域数据描述(rejected transductive inference M-SVDD,RTIM-SVDD)方法,并将该方法应用于机械加工故障诊断当中。首先,RTIM-SVDD通过训练寻求一个尽可能包含所有数据样本的最小超球体作为拒绝检测面,来整体学习样本知识或数据描述,并以一定的拒绝度判别新的测试样本,达到故障检测的目的;其次,应用分别包含各个类别样本的多个超球体,来判别满足一定拒绝度的样本,实现多分类问题。对于模糊样本点归属判别关键问题,本文采用一种新的转导推理规则来进行决策。最后,通过一个仿真实验进行验证,结果证明了RTIM-SVDD的可行性和有效性。
文摘We advance here a novel methodology for robust intelligent biometric information management with inferences and predictions made using randomness and complexity concepts. Intelligence refers to learning, adap- tation, and functionality, and robustness refers to the ability to handle incomplete and/or corrupt adversarial information, on one side, and image and or device variability, on the other side. The proposed methodology is model-free and non-parametric. It draws support from discriminative methods using likelihood ratios to link at the conceptual level biometrics and forensics. It further links, at the modeling and implementation level, the Bayesian framework, statistical learning theory (SLT) using transduction and semi-supervised lea- rning, and Information Theory (IY) using mutual information. The key concepts supporting the proposed methodology are a) local estimation to facilitate learning and prediction using both labeled and unlabeled data;b) similarity metrics using regularity of patterns, randomness deficiency, and Kolmogorov complexity (similar to MDL) using strangeness/typicality and ranking p-values;and c) the Cover – Hart theorem on the asymptotical performance of k-nearest neighbors approaching the optimal Bayes error. Several topics on biometric inference and prediction related to 1) multi-level and multi-layer data fusion including quality and multi-modal biometrics;2) score normalization and revision theory;3) face selection and tracking;and 4) identity management, are described here using an integrated approach that includes transduction and boosting for ranking and sequential fusion/aggregation, respectively, on one side, and active learning and change/ outlier/intrusion detection realized using information gain and martingale, respectively, on the other side. The methodology proposed can be mapped to additional types of information beyond biometrics.
