Fraud is one of the major issues in today’s business world. It was found that 8% increase over last year in the cost per dollar of fraud losses and online retailers are the worst hit as fraudulent transactions grew 32.1% in 2015. Nowadays, data analysis has come to rescue organizations from fraud as it allows companies to create fraud risk profiles and then use existing data to identify potential fraudulent activity. This article explores the commonly used data analysis techniques used to search out fraudulent transactions. They are: Pattern Recognition, Outlier Detection, Regression Analysis, Semantic Modeling, Neural Network, and Anomaly Detection. Read more at: http://it.toolbox.com/blogs/itmanagement/how-to-detect-fraud-using-data-analysis-74726