A non-parametric framework for learning how networks grow
This tool accompanies the paper:
R. Patro+ G. Duggal+, E. Sefer, H. Wang, D. Filippova, C. Kingsford. The missing models: a data driven approach for learning how networks grow. To Appear: Proc. 18th Intl. Conference on Knowledge Discovery and Data Mining (KDD) 2012.
- Equal contributions
For the runnable jar (growcode.jar):
- Java 1.6 or greater
For building and editing source:
- Scala 2.9.1 or greater. (http://scala-lang.org)
- Simple Build Tool (SBT)
The 'runexample.sh' script shows how to run a basic learning and program selection procedure using the SBT framework. The learning procedure is associated with a variety of genetic algorithm parameters that can be customized in each example's '.params' file. Please see the EJC documentation (ECJ link) for more information on what these parameters mean.
- Comment source code more formally
- Make the network properties tested in BestProgramProblem.scala selectable via a config file or command line interface