Traveling salesman
Command lineData structuresGenitic algorithms
Unlike the problem in the genetic algorithm phrase matcher the Traveling Salesman Problem (TSP) is a real problem that you might really use a genetic algorithm to solve.
The problem is, given a list of cities and their distances from each other, find the shortest path that passes through each city exactly once before returning to the starting city. It is relatively easy to write a brute-force algorithm to find a solution to the TSP. Unfortunately it only works for very small problems—i.e. with a small number of cities. This is because the TSP is an example of a class of problem where the size of the problem—in terms of the number of solutions you have to explore—grows so much faster than the size of the problem that no known algorithm can solve them in any reasonable amount of time where “reasonable” means “less than the expected lifetime of the universe”.
But while we cannot write a program that always finds the absolute best solutions we can use techniques such as genetic algorithms to find pretty good solutions. As in the genetic phrase matcher, you will need to figure out a way to represent possible solutions for a given set of cities and then evolve better and better solutions by crossing and mutating those potential solutions.
If you want to try this project you should start with genetic algorithm phrase matcher as a warmup. Some of the code you write for that can be reused for this project while other bits while need to be specific to this project.