Original Post
hi there,
I'm using a structured genetic algorithm to solve a basic knapsack problem and have implemented this to somewhat sucess.
The problem I'm having is that the algorithm never seems to find the best (or even second best) solution to the problem unlike the simple genetic algorithm I have written previously.
I have implemented it where the first two bits of the chromosome represent the high level genes and the next 7 represent the actual genes of the chromosome. I do everything as usual (crossover, mutation etc.) on this chromosome but when it comes to evaluating its fitness I asses whether each gene is active or passive based on its parent gene.
Is this all that is involved or do I need to do anything else?
I'm using a structured genetic algorithm to solve a basic knapsack problem and have implemented this to somewhat sucess.
The problem I'm having is that the algorithm never seems to find the best (or even second best) solution to the problem unlike the simple genetic algorithm I have written previously.
I have implemented it where the first two bits of the chromosome represent the high level genes and the next 7 represent the actual genes of the chromosome. I do everything as usual (crossover, mutation etc.) on this chromosome but when it comes to evaluating its fitness I asses whether each gene is active or passive based on its parent gene.
Is this all that is involved or do I need to do anything else?