No user input required here, just sit back and watch it evolve!
This is the same as my project "Evolutionary Learning V2" except that it tries to find the longest, slowest route to the end. Ok. So. This is going to be complicated to explain... Basically, this is a very, VERY simple example of evolutionary machine learning. The goal is for the orange and red ball to find the most efficient route to the red goal at the end. It "learns" how to do this with a basic genetic model. Every set of 2 runs is a "Generation". Each run is considered an "Organism". Each organism is completed via a sequence of jumps. Each jump is imparted with an initial X and Y velocity. What these velocities are is determined by an "X Gene" and a "Y Gene". Put together, these genes form an "X Genome" and a "Y Genome". Each organism has 1 X and 1 Y genome. After each generation is complete, the speed of each organism is recorded. The faster of the 2 organisms is "selected" out and its X and Y genomes are erased and replaced with the genomes from the slower organism. Then, before the next generation starts, 1 gene in each of all 4 genomes (2 in each organism, 2 organisms) have a chance to "mutate", thereby making the jumps controlled by those genes either stronger, weaker, or the same. Then the process repeats.
