learning sine curve through dummy data, contains same range as the sine curve have through deep learning.
Also include Essemble model to learn model on chunks of data.
Generate data polynomial for sine curve
start with simple model to complex draw training and validation eroor to visulaize model complexity.
using the above model, draw validation set and prediction set to see overfit.
then select best model.
http://gerardnico.com/wiki/data_mining/overfitting
http://docs.aws.amazon.com/machine-learning/latest/dg/model-fit-underfitting-vs-overfitting.html
also when model is overfitting, use norms to reduce parameters.
This is github guide for ensemble model https://github.com/h2oai/h2o-tutorials/tree/master/tutorials/ensembles-stacking