Beginning Anomaly Detection Using

Anomaly detection is the process in 

which an advanced algorithm identifies certain data or data patterns to be anomalous. 

Heavily related to anomaly detection are the tasks of outlier detection, noise removal, 

and novelty detection. In this book, you will explore all of these options as they are all 

basically anomaly detection methods.

 Outlier  Detection

Outlier detection is a technique that aims to detect anomalous outliers within a given 

data set. As discussed, three methods that can be applied to this situation are to train 

only on normal data to identify anomalies by a high reconstruction error, to model a 

probability distribution in which anomalies are labeled based on their association with 

really low probabilities, or to train a model to recognize anomalies by teaching it what an 

anomaly looks like and what a normal point looks like.

Regarding the high reconstruction error, think of the model as having trouble 

labeling an anomaly because it is odd compared to all the normal data points that it has 

seen. Just like how the black swan is really different based on your initial assumption that 

all swans are white, the model perceives this anomalous data point as “different” and has 

a harder time interpreting it.