1-probability that an individual Route will mutate 2-Population size of 1 generation (RoutePop) 3-Size of the tournament selection. In the genetic algorithm, how can I select these three elements (2-2-3) to get the optimal solution out of the result, or are there specific equations to determine them in relation to the number of elements I want to find the best solution from Also, if possible, how can I improve the genetic algorithm because I worked on it to find the shortest distance between the two tracks, but it didn't give me the perfect solution so far(The program I am looking for studies the shortest distance to pass between 51 cities)
Asked
Active
Viewed 153 times
1 Answers
0
Selection alone will not improve enough the Genetic Algorithm. For your problem (I supposed that you want to solve the Travelling Salesman Problem) you need better initialization of the solutions and better genetic operators.
You can start by reading Peter Merz's work on that. For instance this paper:
https://www.researchgate.net/publication/3693015_Genetic_local_search_for_the_TSP_New_results
He solved very large instances (thousands of nodes) with his GA.
regards,
Mihai Oltean
- 159
- 1
- 9