Java: Data Science Made Easy

Using a decision tree with a book dataset

For this example, we will use a file called books.arff. It is shown next and uses four features called attributes. The features specify how a book is bound, whether it uses multiple colors, its genre, and a result indicating whether the book was purchased or not. The header section is shown here:

@RELATION book_purchases@ATTRIBUTE Binding {Hardcover, Paperback, Leather}@ATTRIBUTE Multicolor {yes, no}@ATTRIBUTE Genre {fiction, comedy, romance, historical}@ATTRIBUTE Result {Success, Failure}

The data section follows and consists of 13 book entries:

@DATAHardcover,yes,fiction,SuccessHardcover,no,comedy,FailureHardcover,yes,comedy,SuccessLeather,no,comedy,SuccessLeather,yes,historical,SuccessPaperback,yes,fiction,Failure ...