Once you go Bayesian, there is no evasion!
Organised our project into the following folder structure:
- Iterative_Classification [Jihui and YingYing]
- Loopy_BP [Zihan]
- GNN [Keng Hwee]
- app [Initial Streamlit prototyping for future considerations]
- data [data after node pre-processing done by Jonathan and edge pre-processing done by Leonard]
- raw_data [data collected from Coursemology and MOSS outputs done by Jonathan]
- src [preprocessing code provided by team]
- runs [Tensorboard logs from Keng Hwee's GNN training, run
tensorboard --logdir runs
to see them!] - random_walk [Baseline experiment from Jonathan to find influential nodes]
- requirements.txt [Python modules for reproducibility]