Ground up neural network visualisation exploring the role of machine learning in the direct democracy election process, coded from scratch in Processing.
The simulations are composed of two main elements: the voters (row of circles in the bottom), and the candidates (neural networks on top). Each voter has three "issues" or parameters that they are "interested" in. The candidates set their own "interests" at the start of each generation, and the voters will vote whichever candidate most closely aligns with their issues. After each generation, the votes are tallied. The candidates with the most votes have their genomes crossed over, giving birth to a new generation of candidates.