1 results listed
A growing problem that affects the financial sector
throughout time is financial fraud. Numerous approaches have been
devised to tackle this problem, but data collection has shown to be an
efficient means of funding the automated analysis of vast quantities of
complex data. Data collection has also been essential for identifying
credit card fraud in online purchases. Credit card fraud detection is a
data mining problem. It is challenging for two primary reasons: first,
the characteristics of normal and fraudulent activity are constantly
shifting, and second, the credit card fraud data sets are heavily biassed.
This study examines and evaluates the effectiveness of Decision Tree,
Random Forest, XGBoost, and Logistic Regression using highly
skewed credit card fraud data. The project intends to increase financial
security, decrease false positives and negatives, and increase the
accuracy of fraud detection by combining these strategies. The
suggested method strikes a balance between interpretability and
prediction performance. The approach offers a scalable and effective
fraud detection framework that can be integrated into real-world
banking and payment systems, assisting financial institutions in
mitigating fraud risks while maintaining a seamless user experience.
International Conference on Advanced Technologies, Computer Engineering and Science
ICATCES
K. Maharajan
D. Durga Prasad Reddy
G. Kamalakar Reddy
B. Varun Teja