Beginning Anomaly Detection Using

Figure 5-38.  Defining the columns and loading the data set

Figure 5-39.  Notice that there are categorical labels to deal with, and that there 

are a huge number of columns per data entry

Chapter 5   Boltzmann maChines

---

199

As in Chapter 

2

, you only want to focus on HTTP attacks, so let’s filter the data frame 

to only include them (see Figure 

5-40

).

Figure 5-40.  Filtering all the entries to include only HTTP attacks and dropping 

the service column from the data frame

The new output is shown in Figure 

5-41

.

Figure 5-41.  The columns only consist of HTTP attacks. Here you look at the tail 

end of the data frame

As a reminder, df.tail() performs the same function as df.head() but shows the 

entries from the bottom up as opposed to top down. Also, you can pass a parameter in 

the parenthesis to indicate the number of rows you want to see.

You don’t want values that are strings in your data, so you have to use the label 

encoder as in Chapter 

2

 (see Figure 

5-42

).

Chapter 5   Boltzmann maChines

---

200

The new output is shown in Figure 

5-43

.