AIExplainer
Machine Learning Intermediate 2 min read

What is novelty detection?

A technique used in machine learning to identify new or unusual patterns in data

Novelty detection is a method used to find data points that are significantly different from the rest of the data. This can be useful in a variety of applications, such as detecting anomalies in network traffic or identifying new types of customers.

Novelty detection is like a bouncer at a nightclub. The bouncer is trained to recognize the usual crowd, and when someone new and unusual shows up, they flag them for attention. In the same way, novelty detection algorithms are trained on normal data and flag any new or unusual patterns that appear.

A credit card company uses novelty detection to identify unusual patterns of spending that may indicate fraud. For example, if a customer suddenly starts making large purchases in a foreign country, the algorithm may flag this as a novel pattern and alert the company to potential fraud.

Novelty detection is used in a variety of applications, including fraud detection, network security, and recommendation systems. It can be used to identify unusual patterns in data that may indicate a problem or an opportunity.

One common misconception about novelty detection is that it is the same as outlier detection. While both techniques are used to identify unusual patterns in data, novelty detection is specifically focused on identifying new or unusual patterns, whereas outlier detection is focused on identifying data points that are significantly different from the rest of the data.

Novelty detection has its roots in statistical process control and anomaly detection. The technique has been used in a variety of applications, including quality control and network security, and has become increasingly popular in recent years with the rise of machine learning and big data.

anomaly detection outlier detection one-class classification

Three products for different needs — explore what’s relevant to you.