论文信息 - An Efficient Algorithm for Computing Kernel Function Defined with Anti-unification

An Efficient Algorithm for Computing Kernel Function Defined with Anti-unification

In this paper, we give an algorithm for computing the value of the kernel function K TERM , which takes a pair of terms in first-order logic as its inputs, and facilitates Support Vector Machines classifying terms in a higher dimension space. The value of K TERM (s,t) is given as the total number of terms which subsume both sand t. The algorithm presented in the paper computes K TERM (s,t) without enumerating all such terms. We also implement the algorithm and present some experimental examples of classification of first-order terms with K TERM . Furthermore, we also propose the concept of intentional kernels as a generalization of K TERM .

[1] Luc De Raedt,et al. Kernels and Distances for Structured Data , 2008 .

[2] Thorsten Joachims,et al. Making large-scale support vector machine learning practical , 1999 .

[3] Rocco A. Servedio,et al. Efficiency versus Convergence of Boolean Kernels for On-Line Learning Algorithms , 2001, NIPS.

[4] Shan-Hwei Nienhuys-Cheng,et al. Foundations of Inductive Logic Programming , 1997, Lecture Notes in Computer Science.

[5] Ken Sadohara,et al. Learning of Boolean Functions Using Support Vector Machines , 2001, ALT.

[6] Gordon Plotkin,et al. A Note on Inductive Generalization , 2008 .

[7] Ken Sadohara. On a capacity control using Boolean kernels for the learning of Boolean functions , 2002, 2002 IEEE International Conference on Data Mining, 2002. Proceedings..

[8] Nello Cristianini,et al. An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[9] Stephen Muggleton,et al. Support Vector Inductive Logic Programming , 2005, Discovery Science.

[10] Thomas Gärtner,et al. A survey of kernels for structured data , 2003, SKDD.

[11] David Haussler,et al. Convolution kernels on discrete structures , 1999 .