Data treatment with SPSS

There are always many common data and feature issues to work with for any machine learning project, including this movie recommendation project for which we can use SPSS Modeler.

In comparison with other projects in this book, the data structure here is relatively simple; however, one special issue for the data to be used for this project is about missing values because some users do not rate some movies. To deal with this, SPSS Modeler has a few super nodes to deal with the issue. In other words, we need to develop a special SPSS Modeler Stream, which include nodes for missing value treatments. After this job, we need to separate the data into parts to train and test.

Missing data nodes on SPSS modeler

To deal with missing values and build a special data stream, we need to start with a Type Node with some Super Nodes to handle missing values to be filled with imputed values.

Specifically, you can do this from the Data Audit report, which allows you to specify options...