"MLPClassifier can only process categorial labels (like 0, 1, 2, 3...). But your labels (i.e. Y) are real numbers, so there are illegal. I notice that your task is regression rather than classification. sklearn.neural_network.MLPRegressor may be more suitable, I guess. – Lizhi Liu yesterday"

Thanks for the advice, this took care of the issue. I had to adjust the code a bit more, but it is all squared away.