40 Algorithms Every Programmer Should Know

Anomaly-detection algorithms

The dictionary definition of an anomaly is something that is different, abnormal, peculiar, or not easily classified. It is a deviation from the common rule. In the context of data science, an anomaly is a data point that deviates a lot from the expected pattern. Techniques to find such data points are called anomaly-detection techniques.

Now let's see some applications of anomaly-detection algorithms:

• Credit card fraud

• Finding a malignant tumor in a magnetic resonance imaging (MRI) scan

• Fault prevention in clusters

• Impersonation in exams

• Accidents on a highway

In the upcoming sections, we will see various anomaly-detection techniques.