Beginning Anomaly Detection Using

Figure 5-22.  Looking at the values for the column Amount to see how they were 

standardized

Figure 5-23.  Looking at the values for the column Time to see how they were 

standardized

Chapter 5   Boltzmann maChines

---

190

Awesome; looking much better. Now, you can define your training and testing data 

sets (see Figure 

5-24

).

Figure 5-24.  This is a different process than usual because of how the RBM model 

expects the input

Figure 5-25.  The output shapes of the training and testing sets

You should see something like Figure 

5-25

 as the output.

Chapter 5   Boltzmann maChines

---

191

Getting to the model itself, use the code in Figure 

5-26

.

Figure 5-26.  Initializing the model with a set of parameters

The parameters are as follows:

• 

num_visible: The number of nodes in the visible layer

• 

num_hidden: The number of nodes in the hidden layer

•