Clone this wiki locally
It seems all programmers need to prove they can program their way out of a paper bag. This is my attempt.
I have some simple ideas, an ant colony optimisation, and introduce a level of indirection: putting a maze in a bag and escaping that instead. Not a FizzBuzz in sight.
I gave a talk at #accu2013 describing my approach - you had to be there. Slides To conclude ant colony optimisation under "Which ML?" you need to send the data in the previous table, "Target for See5" through See5. The free demo version is happy with this small dataset. I shall be writing up a genetic algorithm version for ACCU's Overload. My certificate has been added to the conference schedule.