Types of Data Anomalies
Do you know what type of anomalies are you dealing with?
Anomalies are simply anything that doesn’t belong to a well-known group or a pattern and stands out.
In this articles, we will discuss different types of anomalies that we could encounter in handling data or modeling the data and then different detection methods currently in practice.
The technique of identifying rare events or observations that could raise suspicions by being statistically different from the rest of the observations is known as Anomaly Detection (AD).
There are just three types of anomalies:
- Point anomaly,
- Contextual anomaly, and
- Collective anomaly
Point Anomaly:
When a single data point (or datum) or an observation in the data set is far off from the rest of the data, then it’s said to be a point anomaly. They represent an extremum, irregularity, or deviation that occurs randomly with no association with the common pattern in the data.
A point anomaly is also known as global outlier as it is significantly different from the rest of the data set.
