Code associated to the paper: LSTM Networks for Data-Aware Remaining Time Prediction of Business Process Instances, Nicolò Navarin, Beatrice Vincenzi, Mirko Polato and Alessandro Sperduti.
An extension of the popular machine learning library, that implements several graph kernels.
This software is a fork from EDeN, an open source tool for the fats computation of several kernels for graphs.
The DD kernel proposed in the paper A tree-based kernel for graphs is included in the original framework. This version extends the original software with another kernel from the DD family.
The datasets adopted in the papers.
Some other publicly available datasets.
DD kernel with feature selection
This software is related to the paper A memory-efficient graph kernel and allows for the computation of DD kernel with several feature selection criterea.
Online learning on streams of graphs
This software implements the LCB-PA algorithm presented in A Lossy Counting Based Approach for Learning on Streams of Graphs on a Budget. This is a very fast algorithm for online learning on streams of graphs.
Software to transform a graph dataset in a tree dataset, using the features generated by DDK kernel.