International audienceKernel methods are one of the main techniques used for learning on tree structured data. Among them, we interest ourselves to the subtree kernel, which has the combinatorial advantage of being able to enumerate all objects used in the estimation of the similarity between trees. We introduce concepts of DAG reduction and DAG recompression, which lead us to a performant algorithm for computing the kernel. The ability of dealing with a highly compressed database allows us to introduce a new weight function, which manages to capture information in data where weights in the literature do not.Les méthodes à noyaux sont une des approches permettant l’apprentissage à partir de données arborescentes. Parmi elles, nous nous inté...
In this paper, we propose the distributed tree kernels (DTK) as a novel method to reduce time and sp...
International audienceA classical compression method for trees is to exploit subtree repeats in the ...
International audienceSeveral shape similarity measures, based on shape skeletons, are designed in t...
International audienceKernel methods are one of the main techniques used for learning on tree struct...
Kernel methods have been shown to be very effective for applications requiring the modeling of struc...
Abstract — Kernel methods are effective approaches to the modeling of structured objects in learning...
36 pagesInternational audienceTree data are ubiquitous because they model a large variety of situati...
Tree data appear naturally in many scientific domains. Their intrinsically non-Euclidean nature and ...
Kernel methods are effective approaches to the modeling of structured objects in learning algorithms...
Machine learning comprises a series of techniques for automatic extraction of meaningful information...
Tree data appear naturally in many scientific domains. Their intrinsically non-Euclidean nature and ...
Kernel methods have recently been introduced to solve Natural Language Processing and Text Mining pr...
International audienceBy nature, tree structures frequently present similarities between their sub-p...
Due to the expansion of datas, compression algorithms are now crucial algorithms. We address here th...
In this paper, we propose the distributed tree kernels (DTK) as a novel method to reduce time and sp...
International audienceA classical compression method for trees is to exploit subtree repeats in the ...
International audienceSeveral shape similarity measures, based on shape skeletons, are designed in t...
International audienceKernel methods are one of the main techniques used for learning on tree struct...
Kernel methods have been shown to be very effective for applications requiring the modeling of struc...
Abstract — Kernel methods are effective approaches to the modeling of structured objects in learning...
36 pagesInternational audienceTree data are ubiquitous because they model a large variety of situati...
Tree data appear naturally in many scientific domains. Their intrinsically non-Euclidean nature and ...
Kernel methods are effective approaches to the modeling of structured objects in learning algorithms...
Machine learning comprises a series of techniques for automatic extraction of meaningful information...
Tree data appear naturally in many scientific domains. Their intrinsically non-Euclidean nature and ...
Kernel methods have recently been introduced to solve Natural Language Processing and Text Mining pr...
International audienceBy nature, tree structures frequently present similarities between their sub-p...
Due to the expansion of datas, compression algorithms are now crucial algorithms. We address here th...
In this paper, we propose the distributed tree kernels (DTK) as a novel method to reduce time and sp...
International audienceA classical compression method for trees is to exploit subtree repeats in the ...
International audienceSeveral shape similarity measures, based on shape skeletons, are designed in t...