Lab9 Review Raffaele Viola - s309063

[RaffaeleViola]

Your code is well written and structured. The readme file is also pretty clear.

Here are some considerations about your work:

• You could create a class Individual or a similar data structure in order to save the fitness computation of the elite/parent individuals and avoid to recompute it at every step
• A better improvements for the fitness calls would be to create an hash map (python dictionary) and store as a key the genome of the individuals generated (you can use a tuple as key because it is natively hash-able) and as value the fitness. In this way you could avoid to recompute the fitness for already seen individuals.
• You could try to use different type of crossover other than the one cut. For instance you could use the uniform crossover that should be useful with a wide genome like the one of this problem.

Overall good job, the model is standard but effective enough.