Search

This article proposes an algorithm to automatically learn useful transformations of data to improve accuracy in supervised classification tasks. These transformations take the form of a mixture of base transformations and are learned by maximizing the kernel alignment criterion. Because the proposed optimization is nonconvex, a semidefinite relaxation is derived to find an approximate global solution. This new convex algorithm learns kernels made up of a matrix mixture of transformations. This formulation yields a simpler optimization while achieving comparable or improved accuracies to previous transformation learning algorithms based on maximizing the margin. Remarkably, the new optimization problem does not slow down with the availability of additional data allowing it to scale to large datasets. One application of this method is learning monotonic transformations constructed from a base set of truncated ramp functions. These monotonic transformations permit a nonlinear filtering of the input to the classifier. The effectiveness of the method is demonstrated on synthetic data, text data and image data.

The k Nearest Neighbor classifier has been applied to the identification of cancer samples using the gene expression profiles with encouraging results. However, the performance of k -NN depends strongly on the distance considered to evaluate the sample proximities. Besides, the choice of a good dissimilarity is a difficult task and depends on the problem at hand. In this paper, we learn a linear combination of dissimilarities using a regularized version of the kernel alignment algorithm. The error function can be optimized using a semi-definite programming approach and incorporates a term that penalizes the complexity of the family of distances avoiding overfitting. The method proposed has been applied to the challenging problem of cancer identification using the gene expression profiles. Kernel alignment k -NN outperforms other metric learning strategies and improves the classical k -NN based on a single dissimilarity.

This paper proposes to learn the relevant features of remote sensing images for automatic spatio-spectral classification with the automatic optimization of multiple kernels. The method consists of building dedicated kernels for different sets of bands, contextual or textural features. The optimal linear combination of kernels is optimized through gradient descent on the support vector machine (SVM) objective function. Since a na¨ive implementation is computationally demanding, we propose an efficient model selection procedure based on kernel alignment. The result is a weight — learned from the data — for each kernel where both relevant and meaningless image features emerge after training. Excellent results are observed in both multi and hyperspec-tral image classification, improving standard SVM and other spatio-spectral formulations.

International audience; Kernel-target alignment is commonly used to predict the behavior of any given reproducing kernel in a classification context, without training any kernel machine. However, a poor position of the data in feature space can drastically reduce the value of the alignment. This implies that, in a kernel selection setting, the best kernel in a given collection may be associated with a low value of alignment. In this paper, we present a new algorithm for maximizing the alignment by data translation in feature space. The aim is to reduce the biais introduced by the translation non-invariance of this criterion. Experimental results on multi-dimensional benchmarks show the effectiveness of our approach.

Large amounts of histology images are captured and archived in pathology departments due to the ever expanding use of digital microscopy. The ability to manage and access these collections of digital images is regarded as a key component of next generation medical imaging systems. This paper addresses the problem of retrieving histopathology images from a large collection using an example image as query. The proposed approach automatically annotates the images in the collection, as well as the query images, with high-level semantic concepts. This semantic representation delivers an improved retrieval performance providing more meaningful results. We model the problem of automatic image annotation using kernel methods, resulting in a unified framework that includes: (1) multiple features for image representation, (2) a feature integration and selection mechanism (3) and an automatic semantic image annotation strategy. An extensive experimental evaluation demonstrated the effectiveness of the proposed framework to build meaningful image representations for learning and useful semantic annotations for image retrieval.