Open AccessFinancial Fraud Detection Mechanisms to Overcome Trust Issues within Trade Segments
Author(s) -
. Radhika,
Amit Chhabra
Publication year - 2020
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.g5651.059720
Subject(s) - computer science , credit card fraud , similarity (geometry) , support vector machine , field (mathematics) , random forest , credit card , data mining , binary number , matlab , binary classification , financial fraud , machine learning , artificial intelligence , accounting , business , image (mathematics) , mathematics , arithmetic , world wide web , pure mathematics , payment , operating system
Frauds in modern era are cause of concern in almost every field of life. Credit card, money laundering and bank frauds are common and technology has to play important part in overcoming this issue. This paper provides insight into financial frauds leading from malicious users in trading network. To this end several techniques are researched over. To start with price based fraud detection is discussed and then similarity matrix, linear binary patterns, support vector machine and random forest in the field of fraud detection are elaborated. This paper highlights pros and cons of each of such techniques. Dataset required determining classification accuracy of these approaches is synthetically driven. Execution time while determining frauds is critical entity and similarity matrix approach is fast and accurate as compared to random forest, support vector and linear binary patterns. Parameters: Classification Accuracy, Execution time Implementation tool: Matlab 2018 Achievement: support vector machine results are closer to similarity matrix based approach in terms of classification accuracy but execution time of similarity based approach is much less and hence this algorithm is considered better in determining financial frauds.