Fraud Detection Using Data Mining Techniques
Journal Title: Engineering and Scientific International Journal - Year 2017, Vol 4, Issue 4
Abstract
The purpose of this study is to develop a data mining model for fraud detection in various fields such as online transactions (i.e) net banking, tax payers, credit cards etc. and securing data from the third parties(intruders). For fraud detection, very large volumes of data are to be mined and needs some complex mechanism in order to extract full information about those integrators and frauds. It is truly based on baseline study. Some base line studies might help in establishing the extent of fraud detection. These baseline studies uses some current technologies such as ID3 algorithm, Hashing algorithms, Apriori algorithm, query outlier analyses, use case diagrams, pattern reorganization, etc. In order to detect fraud, they depend upon some random audits, informants and some undercover operations. To overcome this problem we used a hashing association mining technique. This study might increase the efficiency and accuracy in fraud identification. For fraud detection, there is no need of any electronically signatures and ensures that some process would not be rejected so that it would increase the performance or working of false alarm rate for fraud detection. Not only have credit cards, but also in every fields such as money laundering, taxed payers and so on. The algorithm is based on the traversal path using hashing approach which is a theoretical approach. Not only managing fraud detection in tax payers but also it is applicable for integrators in network areas in changing data. It uses various tools in finding the frauds and also the integrators. This paper is mainly concentrates on fraud detection and integrators. There is a major need to develop the hierarchical tree structure into different constraints for the fraud detection system. The rate efficiency of this system will determine by analysing raw data and compare with previous techniques. This algorithm will build or improve the alertness by alarm rate which may be classified into two types, one is false alarm rate and the other is actual alarm rate. With the help of these alarm rates, frauds and as well as integrators might be easily detected. So there is a necessity of alarm or alertness for the frauds and integrators.
Authors and Affiliations
Henrik E
Keywords
Related Articles
- EP ID EP631410
- DOI -
- Views 133
- Downloads 0
Finite Element Analysis of Reinforced Glass Beams using ABAQUS Software
Glass is used in buildings as partition wall, window panels, etc in the past centuries as a non structural element which do not carry loads other than its self weight. But it is used as structural element for the past fe...
Analysis of Different Domains of Artificial Intelligence
Artificial Intelligence (AI) is simply the ability of a Machine to act as “intelligence” as a human. It can be as directly showing human intelligence as identifying picture, understanding voices or observing the real-wor...
A novel approach for Liver disorder Classification using Data Mining Techniques
Data mining is an integrated platform for all other soft computing techniques and which is used to identify the expected or probable values from a large storage by using computational algorithms. This paper describes the...
Biometrics Authentication using Electrocardiogram Approach
Biometrics is a secure alternative for traditional methods in identity verification [1]. Electrocardiogram (ECG) is an emerging biometric security mechanism. Biometric measures are used in many different areas and indust...
Online Proffering Approach with Secure Manner using Scrambled Technique
This paper proposes the secured cloud proffering with scrambled technique and uses the Hardware Security Module (temporary storage) to store automatic generating keys for encryption, decryption, and authentication which...