Links:

1. Proposal
2. Report

1. Fix MRAS-Categorical-Subset reproducibility
2. Tune MRAS-Dirchlet-Subset

Use python 3.6. pip install -r requirements.txt

To run:

export PYTHONPATH=.
python sim.py

• UCB
• Thomson Sampling
• Aysm-UCB
• KL-UCB (Needs to be sped up)

We try the following parameter distributions:

1. Categorical
2. Dirchlet
3. Gaussian

We also experiment with the following:

1. Increasing function H.
2. Exploitation param lambda.
3. Simulation allocation M_k.
4. Population size N_o

• Add under bandits folder.
• Add a unit test to verify its working.
• Import it under sim.py

We use sacred for configs and capturing arguments. Read more here. Modify base-config.yaml for more games.

If you need very different args, create a new config file and run as: python sim.py with configs/new-config.yaml

1. Is regret being computed correctly? Right now we are accumulating (best_mean - reward). This could be negative, but averaged over experiments is positive.
2. Is the UCB implementation correct?