The generally accepted process for preparing or tuning a data file for analysis in Watson is shown in the following figure:
The process for preparing a data file in Watson is described in the following steps:
Steps 2 and 3 can be performed solely using Watson. However, you may want to consider extending the process by using a tool that may offer specific data preparation functionalities, such as the IBM SPSS modeler.
For example, earlier in this section, we saw that the Payment Method column contained missing values in 3% of the records in the file. As it turns out, the IBM SPSS modeler is built to recognize several types of missing values: