R for Data Science - 01 - Bank Marketing Poor Model

May 14, 2019
Bank Marketing predictive model that doesn't perform well
Sometimes, when dealing with unbalanced data sets, if you just look at the ROC curve, you can incur in wrong deductions. In this case, even checking the Precision-Recall curve will make us account for the low performance of the model. You can find a better model for the same problem here: [https://gallery.azure.ai/Experiment/R-for-Data-Science-02-Bank-Marketing-Undersampling][1] [1]: https://gallery.azure.ai/Experiment/R-for-Data-Science-02-Bank-Marketing-Undersampling