Fraud Detectionusing Supervised, Unsupervised and Social Network Methods
Business Knowledge Series course
Presented by Presented by Bart Baesens, Ph.D. or Christophe Mues, Ph.D., Assistant Professors at the School of Management of the University of Southampton (UK)
It is estimated that a typical organisation loses about 5% of its revenues due to fraud. Using state of the art analytics, it becomes possible to both detect and prevent fraud. In this course, we will start by first discussing different types of fraud and the challenges to detect it from an analytical perspective. We will then discuss how supervised learning can be used to build fraud detection models using a set of historical labelled observations (e.g. insurance claims, credit card transactions, …). A next section will cover unsupervised learning for anomaly detection. The course concludes by discussing how concepts from social network learning can be used for fraud detection. The course aims at providing a sound mix of both theoretical and technical insights as well as practical implementation details, illustrated by several real-life cases. The course will be highly interactively organised. Background material (selected papers, tutorials and guidelines) will also be provided.
Who should attendAnyone who is involved in fraud issues, or is responsible for fraud detection (all industries).
Before attending this course you should have knowledge in statistics, including descriptive statistics, confidence intervals and hypthesis testing.
This course addresses SAS Enterprise Miner software.
This course addresses SAS Enterprise Miner Software, Base SAS and SAS Social Network Analysis will also be used in this course.
Introduction to Fraud