Has anyone here worked with SharpNeat before? I wrote up a quick experiment and I have it running but I do not see where or how to train the genomes (?, I get caught on terminology as well). From what I can tell it just goes through the process of adding nodes and weights to the genomes. The fitness never really improves. I have some experience using Encog for training ANN's where you create your dataset and feed it into the trainer. I don't see where the process of training the actual ANN's is in SharpNeat.
Can someone tell me where that is or what its called? Or am I just fundamentally missing how NEAT works? In my code I implemented IPhenomeEvaluator<IBlackBox> and INeatExperiment to get the code running.
[link][comment]