Detecting Fraud in Credit Card Transactions u...

Detecting Fraud in Credit Card Transactions using Hidden Markov Model and Fuzzy Logic Neural Networks

Author Name : K. Vivek, K.Snehith, B. Shireeha, B. Rishitha, Ms. P. Swapna

DOI: https://doi.org/10.56025/IJARESM.2024.1205242071

ABSTRACT The use of credit cards has significantly increased as a result of the rapid development of e-commerce and online banking, resulting in a significant number of instances of fraud. In this paper, we propose a novel method for detecting credit card fraud that involves three phases of detection. The initial user authentication and card details verification are carried out in the first phase. In the event that the check is effectively gone through, the exchange is passed to the following stage where Fuzzy c-means grouping calculation is applied to figure out the typical use examples of Mastercard clients in light of their past movement. A doubt score is determined by the degree of deviation from the typical examples and in this way the exchange is named genuine or dubious or false. A neural network-based learning mechanism is used to determine whether a transaction was actually a fraudulent activity or an occasional deviation by a genuine user once it is discovered to be suspicious. Broad trial and error with stochastic models shows that the consolidated utilization of bunching method alongside learning helps in distinguishing fake exercises successfully while limiting the age of misleading problems. Since consumers and financial institutions alike face a significant threat from credit card fraud, sophisticated and robust fraud detection systems must be developed. This examination proposes a creative way to deal with improve the exactness of charge card extortion discovery by consolidating Fuzzy Markov Model and Naive Bayes calculation. . The Hidden Markov Model is a restricted arrangement of states, every one of which is related with a likelihood circulation. Changes between states are constrained by a bunch of conceivable outcomes called progress probabilities.