Beginning Anomaly Detection Using

 Anomaly Detection with the RBM - KDDCUP Data Set

Remember the KDDCUP data set you looked at in Chapter 

2

? Let’s try to apply the RBM 

to it as well. The application will be a similar procedure to that in the previous example, 

but instead of dealing with excessively large values in the data set, you will learn how to 

deal with data that is comprised of a hefty number of zero entries.

Again, you begin by importing all of the necessary modules (see Figure 

5-37

).

Figure 5-37.  Importing the necessary modules

Next, you need to import your data set. Since you’ve used it before, you don’t have to 

do df.head() or print out the shape, but it still helps to get a sense of what the data set 

looks like (see Figure 

5-38

).

Chapter 5   Boltzmann maChines

---

198

The output is shown in Figure 

5-39

.