Beginning Anomaly Detection Using

Figure 4-59.  Code to import packages in Jupyter

Chapter 4   autoenCoders

---

167

You will use the example of credit card data to detect whether a transaction is 

normal/expected or abnormal/anomaly. Figure 

4-61

 shows the data being loaded into a 

Pandas dataframe.

Figure 4-60.  Code to visualize the results

Chapter 4   autoenCoders

---

168

You will collect 20k normal and 400 abnormal records. You can pick different ratios 

to try, but in general more normal data examples are better because you want to teach 

your autoencoder what normal data looks like. Too much of abnormal data in training 

will train the autoencoder to learn that the anomalies are actually normal, which goes 

against your goal. Figure 

4-62

 shows the code to take the majority of normal data records 

with a few abnormal records.