A group project in the context of the CDT in Biomedical AI, at the University of Edinburgh.
Authors: Aryo Pradipta Gema, Dominik Grabarzcyk, Wolf De Wulf
Supervisors: Dr. Javier Alfaro, Dr. Pasquale Minervini, Dr. Antonio Vergari, Dr. Ajitha Rajan
Recompile AnyBURL using scripts/recompile_anyburl.sh
Data used is taken from biokge to obtain the same splits. The train.csv, valid.csv and test.csv can be used directly.
Configuration files can be found ing the config/ folder. An example using default settings for learning and inferrence can be found in config/base
To train one can update the scripts/train.sh file. Inference should loop over the provided time intervals.
Evaulation can be done using the scripts/evaluate.sh file, where paths should be updated to reflect local paths.
Feel free to contact any of the authors via email if you have questions.