Solution to Reply Challenge Optimization Problem on March 12 2020 using Genetic Algorithm
- let we number all available cells for Developer and Managers on map (like [1, 5, 7, 8] for devs, [4, 3, 1] for mans)
- let represent one seating (like dev1 on 1, dev5 sits on 5 cell, man1 seats on cell 17 etc.) as Individual
- then we can create Population from it (take 1000 random seatings for first population), apply crossovers (like create new seating for population: take dev1 seating at cell 5, dev2 seating on cell 6 from one one seating, but man1 and dev3 ceating at cell 7/8 (for example) from other population/and mutations (randomly change seating dev1 to non seating dev2 etc)
- choose N best seatings and then evolve population again, do it until for example 1000 epochs were processed
- best seating at the end will be an answer
Sorry for VERY DISGUSTING code, I wrote it all by myself in less than 4 hours (but this was contest for teams).
I had no time to submit it, but at least let this code remind me
about me having a good time coding this sh*t