Drastic increase in E-commerce has lead to an evolution of credit cards where they act as the navigators to an evolving environment. However, this dramatic increase has also resulted in frauds in credit card transactions. Furthermore, over decades, the technologies have changed, developed and evolved dramatically to give entirely a new face to such frauds. In reality, fraudulent transactions are often amalgamated with genuine transactions. Numerous techniques have been designed to detect and prevent such frauds. Few such approaches include Hidden markov model, Neural network, Decision Tree and so forth. However, these techniques fail to decrease the number of false alerts where in cases of fraud are intimated to the users even when no frauds have actually occurred. In this paper, we present a technique to detect frauds using a combination of one of the novel approach called the Genetic Algorithm and Data mining. Genetic algorithm aims at providing an optimal solution to the problems of fraudulent transactions and also attempts to decrease the number of false alerts while data mining attempts to refine the result obtained from the genetic algorithm.
