AI, genetic algorithms and practical applications

Question

Algorithms

Good afternoon!
On the way of self-education, I came across genetic algorithms and became very interested in the possibility of applying them in practice. Maybe one of the habrausers created and put them into practice? In educational work, with visualization (as there were mutating bicycles or clocks that slipped through on Habré). If you did an imitation of the life of living beings, for example ants, and share your thoughts, ideas, advice or implementation - I will be very grateful! Thank you!

Answer 1

I met on the course of artificial intelligence at the university. At first I thought it was a cool thing - I even tried to simulate evolution - I wrote simple programs. Then I realized that this topic is actually one of the optimization methods. There is an objective function, and input parameters that change at each iteration according to "genetic" laws due to crossing and mutations.
Now a not very close acquaintance, I heard that she is writing a diploma in cryptography using genetic algorithms - so far I have not seen the result.
If you get something interesting - don't forget to write an article on Habr;)

Answer 2

As it was already noted here, genetic algorithms are just a special case of solving an optimization problem. GA is a fairly simple method for the numerical selection of optimal solutions, but often not the most effective. Read more about gradient descent and simulated annealing.
At one time he used GA in his dissertation, but then he abandoned it in favor of analytical solutions.

Answer 3

Used for:
1. Solving a tricky version of the traveling salesman problem.
2. Optimization of schedules.
3. Selection of optimal parameters based on mat. models for some tricky processes.
Simulating the life of living beings is another.

Answer 4

They can look for approximate solutions to all sorts of NP-hard and NP-complete problems.

Answer 5

suitable for a fairly narrow class of problems.
You can look at NEAT.