Crossover combines parts of two strong candidate solutions to produce a new candidate, helping a genetic algorithm search for better solutions. section 2.5
A typical crossover cycle works as follows:
- Evaluate candidate solutions using measures such as delivery time and cost. section 2.5
- Select better-performing candidates and combine useful parts from them. section 2.5
- Apply mutation occasionally to introduce random alternatives, then repeat across generations until improvement stops or a cycle limit is reached. section 2.5
Supply-chain planning
One candidate can contribute a faster delivery route while another contributes a better resource allocation. Combining those parts may produce a more efficient distribution plan. section 2.5
Figure 2.2 Flowchart of a genetic algorithm
Use the genetic-algorithm flowchart to place crossover within evaluation, selection, mutation, and repeated population improvement. section 2.5
