My solution of travelling salesman problem using simulated annealing method.

The program inputs a file in the format: in the first line, enter number n - the number of cities, then in the n lines - the coordinates of the cities written as two numbers. (see data folder for more examples).

Next, the algorithm tries to reduce the initial route length (which is chosen randomly). The difference from gradient descent is that the system has a certain "temperature", which determines the probability of going to a less optimal state with a longer route length (This is necessary so that the algorithm can "get out" of the local minimum with some probability).

As a result, the algorithm outputs two numbers - the total length it came to (if it converged, or if a given number of iterations was passed) and the minimum length for all time (in case the algorithm from the" smaller "local minimum turned out to be" larger")