Filters








196,155 Hits in 2.8 sec

Sparse-Representation-Based Classification with Structure-Preserving Dimension Reduction

Jin Xu, Guang Yang, Yafeng Yin, Hong Man, Haibo He
2014 Cognitive Computation  
MKLSRC using non linear kernel with kernel trick where L1 minimization is used to find sparse representation.  ...  The main objective of MKL-SRC is to classify images in environment having occlusion and noise.  ...  Training data normalized and to compute residual minimization technique is to be used. Generally L1 minimization technique is used.  ... 
doi:10.1007/s12559-014-9252-5 fatcat:aikldxd35jd7zkygkhhlq5m6za

A Novel Linear-Polynomial Kernel to Construct Support Vector Machines for Speech Recognition

Sonkamble
2011 Journal of Computer Science  
The SVMs using LP kernel function classifies the spoken utterances very efficiently as compared to other kernel functions.  ...  Approach: A novel Linear-Polynomial (LP) Kernel function was used to construct support vector machines to classify the spoken word utterances.  ...  Risk minimization means minimizing the functional from a given training data that is minimizing the optimal parameterization.  ... 
doi:10.3844/jcssp.2011.991.996 fatcat:zqfd56bg7na47a7glkdtuiansu

Simultaneous Kernel Learning and Label Imputation for Pattern Classification with Partially Labeled Data

Minyoung Kim
2017 International Journal of Fuzzy Logic and Intelligent Systems  
The second aims to minimize the dual loss in the support vector machines (SVM) classifier learning with respect to the kernel parameters and the missing labels.  ...  This idea of kernel learning has been studied considerably in machine learning and pattern recognition.  ...  The second approach directly minimizes the loss of the kernel classifier based on the underlying kernel.  ... 
doi:10.5391/ijfis.2017.17.1.10 fatcat:hlpjgf57crh6hlhy5ybaz6p5ua

10.1162/153244303765208403

2000 Applied Physics Letters  
A finite concave minimization algorithm is proposed for constructing kernel classifiers that use a minimal number of data points both in generating and characterizing a classifier.  ...  support vector machine classifier that depends on many more data points.  ...  algorithm for a minimal kernel classifier.  ... 
doi:10.1162/153244303765208403 fatcat:mk222ynbozgfxoqsjf4wk4uzzi

Research on the Optimal Information Retrieval Based on SVM

Jianming Sun, Qingli Sun
2014 International Journal of Control and Automation  
Then to classify the information with the sentence sequences function and optimize the distribution of time retrieval. This system is proved to have excellent performance of classifying and retrieval.  ...  The number of correct classified positive examples (TP), the number of wrong classified positive examples (FP) and wrong classified negative examples (FN) are calculated by the classification results.  ...  Figure 1 . 1 Structural Risk Minimization Principle Figure 2 . 2 LIBSVM This paper only predicts custom kernel function and ignores other support such as oneclaSSSVM, SVR.  ... 
doi:10.14257/ijca.2014.7.8.38 fatcat:acw2bngzlbfidnqipfdp34jja4

Kernel classifier with Correntropy loss

Rosha Pokharel, Jose C. Principe
2012 The 2012 International Joint Conference on Neural Networks (IJCNN)  
, and hence, a better classifier can be obtained.  ...  Since gradient descent can still be used with the C-loss and the kernel mapper, the classifier can be easily trained without performance penalty, compared to the SVM, which makes the approach very practical  ...  A classifier learns a correct way of predicting labels or a decision rule for an incoming data, based on some sort of learning algorithm that focuses on minimizing the empirical risk.  ... 
doi:10.1109/ijcnn.2012.6252721 dblp:conf/ijcnn/PokharelP12 fatcat:4etjnodwsncytcnsx77ishdoy4

Automatic Capacity Tuning of Very Large VC-Dimension Classifiers

Isabelle Guyon, Bernhard E. Boser, Vladimir Vapnik
1992 Neural Information Processing Systems  
The technique is applicable to a wide variety of classifiers, including Perceptrons, polynomial classifiers (sigma-pi unit networks) and Radial Basis Functions.  ...  In this paper we show that even high-order polynomial classifiers in high dimensional spaces can be trained with a small amount of training data and yet generalize better than classifiers with a smaller  ...  The supporting patterns (white) lie on the margin. (1) Polynomial classifier of order two (sigma-pi unit network), with kernel K(x, x') = (x. x' + 1)2. (2) Kernel classifier (RBF) with kernel K(x,x) =  ... 
dblp:conf/nips/GuyonBV92 fatcat:vgken6awjjcexd24ewq25iugvm

A comparative analysis of structural risk minimization by support vector machines and nearest neighbor rule

Bilge Karaçalı, Rajeev Ramanath, Wesley E. Snyder
2004 Pattern Recognition Letters  
In this paper, we present a technique that combines these two classifiers by adopting a nearest neighbor rule-based structural risk minimization classifier.  ...  Using synthetic and real data, the classification technique is shown to be more robust to kernel conditions with a significantly lower computational cost than conventional SVMs.  ...  In terms of providing a consistent classifier accuracy for various choices of kernel parameters, guaranteed convergence and fast computation, the NNSRM method clearly outperforms the SVM classifier.  ... 
doi:10.1016/j.patrec.2003.09.002 fatcat:oupur7ofi5bt7fokuapyzwonuu

Kernel Robust Bias-Aware Prediction under Covariate Shift [article]

Anqi Liu, Rizal Fathony, Brian D. Ziebart
2017 arXiv   pre-print
To overcome this issue, we extend the representer theorem to the RBA setting, enabling minimization of regularized expected target risk by a reweighted kernel expectation under the source distribution.  ...  By applying kernel methods, we establish consistency guarantees and demonstrate better performance of the RBA classifier than competing methods on synthetically biased UCI datasets as well as datasets  ...  Methods We evaluate our approach and five other methods: Kernel robust bias aware classifier (Kernel Robust) adversarially minimizes the target distribution logloss using kernel methods, trained using  ... 
arXiv:1712.10050v1 fatcat:g5xhvjrh4bajfbxzneqg7id5rm

Nonlinear L1-norm minimization learning for human detection

Ran Xu, Jianbin Jiao, Qixiang Ye
2011 2011 18th IEEE International Conference on Image Processing  
The NL-LML method first maps the samples into a space based on the kernel function, and then combines the reformulated samples in the transformed space with the LML model to learn a classifier.  ...  Inspired by the successful applications of L1-norm minimization learning (LML) for human detection, we propose a new nonlinear L1-norm minimization learning method (NL-LML).  ...  Furthermore, previous work has shown that the LML model pursues the VC-dimension minimization and further guarantees minimization of the upper bound on test error [12] [13] .  ... 
doi:10.1109/icip.2011.6116488 dblp:conf/icip/XuJY11 fatcat:aodmnzqd4vbjpbaly2kl5cb3uy

Kernel Classification via Integrated Squared Error

JooSeuk Kim, Clayton D. Scott
2007 2007 IEEE/SP 14th Workshop on Statistical Signal Processing  
Our classifier is sparse, like SVMs, and performs comparably to state-of-the-art kernel methods.  ...  Index Terms-kernel methods, integrated squared error, sparse classifiers, quadratic programming, difference of densities  ...  A kernel classifier has the form g(x) = sign n i=1 α i y i k(x, x i ) , where α i are parameters and k is a kernel function, often taken to be a Gaussian kernel.  ... 
doi:10.1109/ssp.2007.4301366 fatcat:6mdvovl4jnb5ra5il35jaq2l2m

Signal discrimination using a support vector machine for genetic syndrome diagnosis

A. David, B. Lerner
2004 Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004.  
The study implements the SVM structural risk minimization concept in searching for the optimal setting of the classifier kernel and parameters.  ...  Results show accurate performance of the SVM in classifying FISH signals in comparison to other state-ofthe-art machine learning classifiers, indicating the potential of an SVM-based genetic diagnosis  ...  Structural risk minimization for the SVM was explored empirically by employing different families of functions, characterized by their kernels, and different settings of classifier parameters.  ... 
doi:10.1109/icpr.2004.1334573 dblp:conf/icpr/DavidL04 fatcat:ttj4quhba5f5fgcj4y35d4cvb4

Automatic model selection for the optimization of SVM kernels

N.E. Ayat, M. Cheriet, C.Y. Suen
2005 Pattern Recognition  
For the former, we also propose a new minimization scheme.  ...  The proposed methodology suggests the use of a new model selection criterion based on the estimation of the probability of error of the SVM classifier.  ...  Once the model selection process is terminated, the classifiers are trained with the optimized kernel parameters on the original training set.We have implemented three variants of the minimization algorithm  ... 
doi:10.1016/j.patcog.2005.03.011 fatcat:xjbgdqnip5amfjvnxcmqixw2f4

Classification of RR-Interval and Blood Pressure Signals Using Support Vector Machine for different Postures

Indu Saini, Arun Khosla, Dilbag Singh
2012 Journal of clean energy technologies  
Therefore support vector machine, a classifier motivated from statistical learning theory, is used here for classifying the subjects based on lying and standing postures.  ...  The efficiency of SVM lies in the choice of the kernel for a given problem.  ...  Kernel-based SVM Classifier The accuracy of an SVM model is largely dependent on the selection of the kernel method applied.  ... 
doi:10.7763/ijcte.2012.v4.490 fatcat:nnt4373ionc3zju2jm2fkolwja

A least square kernel machine with box constraints

Jayanta Basak
2008 Pattern Recognition (ICPR), Proceedings of the International Conference on  
In this paper, we present a least square kernel machine with box constraints (LSKMBC).  ...  Experimentally we validate the performance of the classifier and show that it is able to outperform SVM and LSSVM on certain real-life datasets.  ...  We have shown the relationship of the classifier with the existing least square kernel classifiers such as RLSC and LSSVM.  ... 
doi:10.1109/icpr.2008.4761717 dblp:conf/icpr/Basak08a fatcat:c2pp7ro755b6tgeh3xqqem52cq
« Previous Showing results 1 — 15 out of 196,155 results