A survey of kernels for structured data

@article{Grtner2003ASO,
  title={A survey of kernels for structured data},
  author={Thomas G{\"a}rtner},
  journal={SIGKDD Explor.},
  year={2003},
  volume={5},
  pages={49-58}
}
  • T. Gärtner
  • Published 1 July 2003
  • Computer Science
  • SIGKDD Explor.
Kernel methods in general and support vector machines in particular have been successful in various learning tasks on data represented in a single table. Much 'real-world' data, however, is structured - it has no natural representation in a single table. Usually, to apply kernel methods to 'real-world' data, extensive pre-processing is performed to embed the data into areal vector space and thus in a single table. This survey describes several approaches of defining positive definite kernels on… 
Kernels and Distances for Structured Data
TLDR
A general method for constructing a kernel following the syntactic structure of the data, as defined by its type signature in a higher-order logic, and the main theoretical result is the positive definiteness of any kernel thus defined.
Kernel Methods for Tree Structured Data
TLDR
This thesis proposes to encode the input trees by an algorithm able to project the data onto a lower dimensional space with the property that similar structures are mapped similarly, thus reducing its sparsity with respect to traditional tree kernel functions.
Kernel methods in machine learning
We review machine learning methods employing positive definite kernels. These methods formulate learning and estimation problems in a reproducing kernel Hilbert space (RKHS) of functions defined on
A Short Tour of Kernel Methods for Graphs
Machine learning research has – apart from some exceptions – originally concentrated on learning from data that can naturally be represented in a single table without links between the instances. Due
Kernel Methods for Graphs: A Comprehensive Approach
TLDR
This paper focuses on Kernel Methods that face up to data that can be represented by means of graphs, by providing an in-depth review through a comprehensive approach to the research hints and the main open problems in this area of research.
Fast learning of relational kernels
We develop a general theoretical framework for statistical logical learning with kernels based on dynamic propositionalization, where structure learning corresponds to inferring a suitable kernel on
Kernel Methods for Analyzing Structured-data
TLDR
The idea of the convolution kernel that is a general framework for designing kernels for structured data, and some examples of such kernels are introduced.
Linear Online Learning over Structured Data with Distributed Tree Kernels
TLDR
Distributed Trees (DT) is investigated as an efficient way to use structured data in online learning by embedding the huge feature space of the tree fragments into small vectors, so enabling the use of linear versions of kernel machines over tree structured data.
Developments in kernel design
TLDR
A concise overview of kernels, with a special attention to non-standard or heterogeneous data sources (e.g. non-numerical or structured data) and the world of possibilities that kernel design opens for the principled analysis of special or new application domains is discussed.
Weighted decomposition kernels
TLDR
Experimental evidence is reported that the proposed kernel is highly competitive with respect to more complex state-of-the-art methods on a set of problems in bioinformatics.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 53 REFERENCES
Multi-Instance Kernels
TLDR
A kernel on multi-instance data that can be shown to separate positive and negative sets under natural assumptions is shown and compares favorably with state of the art multi- instance learning algorithms in an empirical study.
Kernels for structured data
TLDR
This paper proposes a kernel for structured data, proves that it is positive definite, and shows how it can be adapted in practical applications.
On Graph Kernels: Hardness Results and Efficient Alternatives
TLDR
As most ‘real-world’ data is structured, research in kernel methods has begun investigating kernels for various kinds of structured data, but only very specific graphs such as trees and strings have been considered.
Kernels for Semi-Structured Data
TLDR
This paper model semi-structured data by labeled ordered trees, and presents kernels for classifying labeled ordering trees based on their tag structures by generalizing the convolution kernel for parse trees introduced by Collins and Duffy (2001).
Diffusion Kernels on Graphs and Other Discrete Input Spaces
TLDR
This paper proposes a general method of constructing natural families of kernels over discrete structures, based on the matrix exponentiation idea, and focuses on generating kernels on graphs, for which a special class of exponential kernels called diffusion kernels are proposed.
Marginalized kernels for biological sequences
TLDR
This work proposes a reasonable way of designing a kernel when objects are generated from latent variable models (e.g., HMM), and derives several marginalized kernels useful for biological sequences.
Convolution Kernels for Natural Language
TLDR
It is shown how a kernel over trees can be applied to parsing using the voted perceptron algorithm, and experimental results on the ATIS corpus of parse trees are given.
An introduction to kernel-based learning algorithms
This paper provides an introduction to support vector machines, kernel Fisher discriminant analysis, and kernel principal component analysis, as examples for successful kernel-based learning methods.
Dynamic Alignment Kernels
There is much current interest in kernel methods for classi cation re gression PCA and other linear methods of data analysis Kernel methods may be particularly valuable for problems in which the
Marginalized Kernels Between Labeled Graphs
TLDR
A new kernel function between two labeled graphs that is based on an infinite dimensional feature space, so it is fundamentally different from other string or tree kernels based on dynamic programming and presents promising empirical results in classification of chemical compounds.
...
1
2
3
4
5
...