Fitness Evaluation

• Normalisation or scaling of fitness values required to prevent good solution overwhelming later generations

• known as “premature convergence”
• similar to the ‘local minima’ problem with neural networks