Search

Artificial Intelligence & Machine Learning

Insurance Fraud Detection

A data scientist approach where the data of the users were analyzed clustered and processed in unsupervised learning methods.

Unsupervised Anomaly Detection Engine

FRAUDULENT BEHAVIOUR IS NOTHING NEW FOR BIG INSURANCE COMPANIES

Fraud detection is a challenging problem. The fact is that fraudulent transactions represent a very small fraction of activity within an organization. The challenge is that a small percentage of activity can quickly turn into big losses without the right tools and systems in place.

Criminals are crafty. As traditional fraud schemes fail to pay off, fraudsters have learned to change their tactics.

The good news is that with advances in machine learning, systems can learn, adapt and uncover emerging patterns for preventing fraud.

CHALLENGE

The customer, a large insurance company, needed to identify potential fraudulent users.

To keep the number of frauds as small as possible it is important to detect frauds right at the beginning. Data Reply faced the challenge to separate fraudulent users from honest ones, so that no righteous person would be suspected to be guilty of an offence.

FOCUS ON PREDICTION & PRESCRIPTION

Data Reply developed an unsupervised anomaly detection engine by data manipulation and feature construction.
The user's data were analyzed and clustered into types of data, so identifying those which were most predictable of insurance fraud. Data Reply had a data scientist approach to process data in unsupervised learning methods.