Hybrid Approach for Fraud Detection in Financial Sector

  • A.Varija Naga Soujanya, Dr. C. Karthikeyan, Ch. Manasa, Y. Narendra Surya

Abstract

Credit card cheat exchanges are turning out to be progressive step by step and it is getting
increasingly hard for people to break down fake exchanges by examining subsequently and it has
gotten vital for people to build up a canny framework to decide fake exchanges. A few keen
calculations can be utilized right now peculiar identification. In this paper we executed Neural
Networks, Decision Tree and Random Forest to figure out which calculation is best fit as far as
time taken, recall, accuracy, precision and exactness. We had the option to define consequences
of 284,407 exchanges over a time of two days in September 2013. We had the option to recognize
that three models are practically equivalent with regards to exactness yet arbitrary backwoods is
increasingly exact. Our goal is to look at models that foresee which exchanges could be fake with
high exactness and propose a hybrid technique that can be more precise than existing ones and
help in easy implementation.

Published
2020-05-01
How to Cite
A.Varija Naga Soujanya, Dr. C. Karthikeyan, Ch. Manasa, Y. Narendra Surya. (2020). Hybrid Approach for Fraud Detection in Financial Sector. International Journal of Advanced Science and Technology, 29(7s), 3038-3046. Retrieved from https://sersc.org/journals/index.php/IJAST/article/view/17379