UU2_gp_hyperparameter_estimation pipline needs to be updated

The current pipeline for the `uu2_gp_hyperparamter_estimation` dataset runs but returns the same predictions for all rows. This is due to only one feature being generated for the problem, and it is the same for all rows. The current pipeline is not using the data in the individual CSV files in the dataset. This pipeline should be updated, or a new primitive should be created, to generate additional features from the data scored in the individual CSV files in this dataset.