A Methodology for Outlier Detection in Financial Transactions
This paper presents a design of identifying abnormal transactions in regular financial transactions. This paper covers about the outliers, causes of outliers and methodologies to identify these outliers. Outliers are defined as a data object deviates significantly from other objects in the set. Basically there are two approaches proposed in this paper. Both approaches use semi-supervised and supervised learning data mining algorithms.
Keywords - Data Mining, Outliers.