Voting Rules via Neural Networks

Automated design of voting rules using neural networks - written in order to satisfy the requirements of the honors program at Oberlin College. Thesis and presentation included.

• Commands to run experiments for 3 candidates and 20 voters in Main_nC3.sh, and 5 candidates and 40-80 voters on Main_nC5.sh.

• To generate preference profiles, run the commands at the top /automated_voting/voting/profiles.py

• To test baselines, run the commands at the top of /automated_voting/voting/election.py

• To test neural network, look into /automated_voting/algorithms/neural_network.py

• To run everything look into main.py

• Main thesis

More details to come soon.