Ed Tobias presents on the topic of data mining. He defines data mining as using automated methods to detect patterns in current and historical data to predict future trends and analyze large amounts of information efficiently. Data mining is used in auditing for risk assessment, control evaluation, and fraud detection. It allows analyzing 100% of transactions rather than a sample. Case studies show data mining identified millions of dollars in exception claims in a health insurance audit and uncovered collusion in government procurement that random sampling would have missed.