• Uses Simulated Annealing from the optim.jl package.
• Uses a very naive cost function.
• See demo.jl for usage example. This demo may need to be run a few times before you get lucky and hit the exact solution.

Disclaimer: I've solved exactly one Sudoku in my life before. I'm not sure I understand the rules. What I understand, I've taken from the Wikipedia article on Sudoku. Also, I'm definitely not trying to solve Sudoku problems in a smart way. I just wanted to show how easy it is to solve seemingly "complex" problems using something as stupid as simulated annealing.

In contrast to the simplistic solution I've written up, there is a very efficient Sudoku solver at https://github.com/IainNZ/MathProg.jl/blob/master/examples/sudoku.jl, which highlights Julia's tools for advanced optimization.