Two partition models are created on the same training dataset predicting the same target. Each uses all available predictor variables. The first model uses 7 splits, and the second model uses 30. On the training dataset, each yields an R^2 value of .75. Typically splitting 30 times would be difficult to handle for a computer, but you have a SICK machine that can crush it no problem. That being the case, the second model (30 splits) will definitely perform better on a new Validation dataset. a. Agree
b. Disagree