Abstract
Fraud identification is a crucial issue facing large economic institutions, which has caused due to the rise in credit card payments. This paper brings a new approach for the predictive identification of credit card payment frauds focused on Isolation Forest and Local Outlier Factor. The suggested solution comprises of the corresponding phases: pre-processing of data-sets, training and sorting, convergence of decisions and analysis of tests. In this article, the behavior characteristics of correct and incorrect transactions are to be taught by two kinds of algorithms local outlier factor and isolation forest. To date, several researchers identified different approaches for identifying and growing such frauds. In this paper we suggest analysis of Isolation Forest and Local Outlier Factor algorithms using python and their comprehensive experimental results. Upon evaluating the dataset, we received Isolation Forest with high accuracy compared to Local Outlier Factor Algorithm
Highlights
Credit cards have been used in people's everyday lives to go shopping
According to the U.S Federal Trade Commission survey, the identity fraud rate stayed steady until the mid-2000s, but throughout 2008 it rose by 21 points.such frauds impact financial institutions and the consumers [1]
The key problems involved in the identification of credit card fraud are: Immense data is collected on a regular basis and the model construct must be quick sufficiently respond to the scandal in time
Summary
Credit cards have been used in people's everyday lives to go shopping. For buying products and services this purchasing can be offline as well as internet. Credit card purchases are defined by tracking the conduct of purchases into two classifications: fraudulent and non-fraudulent. Depending on these two groups correlations are generated and machine learning algorithms are used to identify suspicious transactions. The key problems involved in the identification of credit card fraud are: Immense data is collected on a regular basis and the model construct must be quick sufficiently respond to the scandal in time. The fraudsters used advanced tactics against the system [3] To handle these challenges, we go for the following: The model used would be easy and accurate sufficiently identify the phenomenon and recognize it as a suspicious activity as as possible.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
