Programming Homework Help

The goal of this project is to define an evolutionary computation problem, then run a genetic algorithm to solve the problem you defined, and finally discuss and analyze the results.

The problem setup has two main components:

1. What is your search space?
   • First think of what the search space is conceptually. For example, in the symbolic regression, the search space was “mathematical expressions”. You might also consider things like “2d maze”, “2d videogame level” or “fractal art”.
   • Then think of how you would represent individual candidate solutions in this space as a data structure For mathematical expressions, it can be an expression tree. For this project, it will be easiest if you can represent solutions as lists, arrays, or trees.
2. What does a good solution in this space look like?
   • Again, think first conceptually, although with an eye towards something that is possible to implement. For a maze, it might be “there’s a long path through the maze”.
   • Then turn this into an evaluation function that takes in a candidate individual and scores its fitness. For a maze, it might be the length of the longest path (or -1000 if there is no path).

Once you have a problem set up, experiment with evolving solutions in this space. You can use either your own implementation or an existing evolutionary computation library like DEAP.

Finally, report your results in two forms:

1. Show some examples of the best output that was evolved, and its fitness levels.
2. Plot some graphs showing the progress of the evolutionary algorithm. Vary a few parameters like mutation rate and report how this changes the results

Useful code(optional):

https://colab.research.google.com/drive/1noQfQ9o1a5fZGLFUAJAQqUn-cu1VkeMN

https://colab.research.google.com/drive/1FEyNSp48kIVPjyTBtIcGGnEPH5cG0Qyc

https://colab.research.google.com/drive/1kJcWL93WYLNjfO8Bp-PhddK1IJ9FbdMO