: A major problem which is affecting growth in financial services is “CREDIT CARD FRAUD”. Many organizations
lost their amount due to these frauds. Even though many research studies are made on fraud detection, they lack on analyzing
data extracted from actual transaction, due to privacy issues. Here, certain machine learning algorithms are used to detect
fraudulent transactions.
First, the Standard methods are applied. Then, in combination hybrid methods such as AdaBoost and majority voting are
applied. A publicly available datasets are used so that model efficacy can be evaluated. Later, a real-world credit card data set is
analyzed which is taken from certain financial institution. Further, to estimate the robustness of the algorithm, noise is added to
the samples.
