A memory efficient graph kernel
By Giovanni Da San Martino, Nicoló Navarin, Alessandro Sperduti. 2012
16 citations recorded by the source corpus.
In this paper, we show how learning models generated by a recently introduced state-of-the-art kernel for graphs can be optimized from the point of view of memory occupancy. After a brief description of the kernel, we introduce a novel representation of the explicit feature space of the kernel based on an hash function which allows to reduce the amount of memory needed both during the training phase and to represent the final learned model. Subsequently, we study the application of a feature selection strategy based on the F-score to further reduce the number of features in the final model. On two representative datasets involving binary classification of chemical graphs, we show that it is actually possible to sensibly reduce memory occupancy (up to one order of magnitude) for the final model with a moderate loss in classification performance.
Subfields: Artificial Intelligence, Materials Chemistry
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Christopher Morris
- Nils M. Kriege
- Fredrik D. Johansson
- Petra Mutzel
- Gaurav Rattan
- Mihaela van der Schaar
- William L. Hamilton
- Ahmed M. Alaa
In Materials Chemistry
Citations in the corpus
- A survey on graph kernels (2019, methodology): “Da San Martino et al. (2012b) proposed decomposing a graph into trees and applying a kernel defined on trees.”