Follow these steps to prepare the data in the data preparation stage:
- Get your data and partition it.
- Run a Neural Net model again, with a Random seed of 5000. We will get an accuracy of around 80%.
- We will now create a new model and add one additional predictor. If you go back to the list of top ten predictors list, we will find predictors such as the number of speakers and TVs that customers purchased. Hence, it makes sense to have a predictor that shows the total number of items purchased.
- Go to the Field Ops palette and connect the Partition node to a Derive node:
A Derive node allows us to create additional variables. ...