Multiple kernel learning

About: Multiple kernel learning is a research topic. Over the lifetime, 1630 publications have been published within this topic receiving 56082 citations.

A Combination Model of Robust Principal Component Analysis and Multiple Kernel Learning for Cancer Patient Stratification

Abstract: In recent years, bioinformatics has been significantly contributing to patient stratification that is very crucial for early detection of cancer diseases. In particular, stratification or classification of patients is to divide patients into subgroups that will be offered effective treatment regimens. However, current methods have to face two major challenges in analyzing large biomedical datasets when stratifying cancer patients. Firstly, the datasets are very big with a high number of features. Secondly, because the public data is available and heterogeneous, there is a great need of combining multiple data sources, providing more comprehensive and informative datasets. A variety of methods has been proposed to tackle these challenges, but they have often solved one or the other separately. Handling noisy data encountered another difficulty in data integration. In this paper, we have proposed an efficient model, combining of the robust principal component analysis-based dimensionality reduction and feature extraction with classification based on multiple kernel learning. The proposed method resolved the above-mentioned problems in cancer patient stratification. The model obtained high accuracy with 92.92% and significant statistical tests. These results hold great promise, supporting cancer research, diagnosis, and treatment.

Multiple spectral kernel learning and a gaussian complexity computation

Abstract: Multiple kernel learning MKL partially solves the kernel selection problem in support vector machines and similar classifiers by minimizing the empirical risk over a subset of the linear combination of given kernel matrices. For large sample sets, the size of the kernel matrices becomes a numerical issue. In many cases, the kernel matrix is of low-efficient rank. However, the low-rank property is not efficiently utilized in MKL algorithms. Here, we suggest multiple spectral kernel learning that efficiently uses the low-rank property by finding a kernel matrix from a set of Gram matrices of a few eigenvectors from all given kernel matrices, called a spectral kernel set. We provide a new bound for the gaussian complexity of the proposed kernel set, which depends on both the geometry of the kernel set and the number of Gram matrices. This characterization of the complexity implies that in an MKL setting, adding more kernels may not monotonically increase the complexity, while previous bounds show otherwise.

Decomposition and Extraction: A New Framework for Visual Classification

Abstract: In this paper, we present a novel framework for visual classification based on hierarchical image decomposition and hybrid midlevel feature extraction. Unlike most midlevel feature learning methods, which focus on the process of coding or pooling, we emphasize that the mechanism of image composition also strongly influences the feature extraction. To effectively explore the image content for the feature extraction, we model a multiplicity feature representation mechanism through meaningful hierarchical image decomposition followed by a fusion step. In particularly, we first propose a new hierarchical image decomposition approach in which each image is decomposed into a series of hierarchical semantical components, i.e, the structure and texture images. Then, different feature extraction schemes can be adopted to match the decomposed structure and texture processes in a dissociative manner. Here, two schemes are explored to produce property related feature representations. One is based on a single-stage network over hand-crafted features and the other is based on a multistage network, which can learn features from raw pixels automatically. Finally, those multiple midlevel features are incorporated by solving a multiple kernel learning task. Extensive experiments are conducted on several challenging data sets for visual classification, and experimental results demonstrate the effectiveness of the proposed method.

Applying Multiple Kernel Learning to Automatic Genre Classification

Abstract: In this paper we demonstrate the advantages of multiple-kernel learning in the application to music genre classification. Multiple-kernel learning provides the possibility to adaptively tune the kernel settings to each group of features independently. Our experiments show the improvement of classification performance in comparison to the conventional support vector machine classifier.

Semi-supervised classification of multiple kernels embedding manifold information

Abstract: For semi-supervised learning, we propose the Laplacian embedded multiple kernel regression model. As we incorporate the multiple kernel occasion into manifold regularization framework, the models we proposed are flexible in many kinds of datasets and have a solid theoretical foundation. The proposed model can solve the two problems, which are the computation cost of manifold regularization framework and the difficulty in dealing with multi-source or multi-attribute datasets. Though manifold regularization is a convex optimization formulation, it often leads to dense matrix inversion with computation cost. Laplacian embedded method we adopted can solve the problem, however it lacks the proper ability to process complex datasets. Therefore, we further use multiple kernel learning as a part of the proposed model to strengthen its ability. Experiments on several datasets compared with the state-of-the-art methods show the effectiveness and efficiency of the proposed model.