UPI FRAUD DETECTION USING

MACHINE LEARNING
AIM
 OBJECTIVES

• 1. Investigate the use of machine learning algorithms for fraud

detection in financial transactions.
• 2. Design and develop a real-time monitoring system for
continuous fraud detection and prevention.
• 3. Assessing the performance of the suggested approach in
comparison to conventional rule-based systems.
• 4. Exploring proactive measures for fraud prevention, such as
dynamic risk scoring and adaptive thresholds.
• 5. Analyze scalability and deployment considerations for
implementing the proposed system in real – world financial
institutions
 INTRODUCTION

• For organizations, financial institutions, and people everywhere,

detecting and preventing fraud in financial transactions is a top
priority. The need to investigate more sophisticated techniques has
arisen as sophisticated fraud has made clear the limitations of
conventional rule-based systems. This study explores how real-time
monitoring systems and machine learning algorithms can be used
to improve financial transaction fraud detection and prevention
capabilities.
• Large volumes of transactional data can be automatically mined for
patterns and abnormalities using machine learning algorithms,
leading to more precise and adaptable fraud detection.
• The necessity for a more effective and efficient strategy to
counteract changing fraud strategies is what motivates the use of
machine learning algorithms and real-time monitoring systems.
 PROJECT DESCRIPTION

• The dataset used for the research is a synthetic dataset

generated for the purpose of this study. It contains information
about financial transactions, including transaction IDs,
customer IDs, transaction amounts, transaction timestamps,
regions, states, customer categories, and account balances.
The dataset consists of 10000 records and includes
characteristics such as geographical information, customer
profiles, and transaction details
S.NO PAPER NAME AUTHOR NAME METHODOLOGY FUTURE WORK
1 “A Gonzalo The methodology of this study As for future works,
Comparative LITERATURE REVIEW
involves conducting a practical the study may
Analysis of analysis to evaluate the suggest several
Boost” performance of a novel technique in directions for further
terms of training speed, research. This could
generalization performance, and involve exploring
parameter setup. This likely includes enhancements or
selecting appropriate datasets, modifications to the
preprocessing steps, and novel technique to
performance metrics. improve its
performance.
2 “Credit Card Rahul Goyal, The research paper proposes Future work
Fraud Amit Kumar employing a combination of the suggested in the
Detection in Synthetic Minority Oversampling paper could
Data Mining Technique (SMOTE) followed by the encompass
using XGBoost Boost classification algorithm for exploring different
Classifier” fraud activity classification. aspects.
3 “Customer Yixuan Zhang, The paper presents a fraud As for future work,
Transaction Jialiang Tong detection algorithm based on several avenues
Fraud XGBoost. Initially, data cleaning is could be explored.
Detection conducted to address anomalies. To Researchers may
Using XGBoost tackle the issue of unbalanced label delve into
Model” distribution. optimizing the
parameters of the
S.NO PAPER NAME AUTHOR NAME METHODOLOGY FUTURE WORK
4 “Customer Yixuan Zhang, The methodology employed in this In terms of future
Transaction Jialiang Tong,Ziyi paper for fraud detection using works, there are
Fraud Wang,Fengqiang XGBoost involves a systematic several avenues for
Detection Gao approach to address various improvement and
Using Xgboost challenges in the data and modeling exploration, fine-
Model” process. Recognizing the tuning of
imbalanced distribution of labels, hyperparameters
the Synthetic Minority Oversampling and exploring
Technique (SMOTE) was applied to ensemble methods
oversample the minority class. could lead to
improved model
accuracy and
robustness
5 Influence of Dr. C. Victoria The paper introduced a For future work,
Optimizing Priscilla, D. methodology aimed at enhancing researchers could
XGBoost to Padma Prabha the XGBoost model's capability to explore extending
handle Class handle class imbalance in datasets these optimization
Imbalance in without resorting to traditional techniques to other
Credit Card sampling techniques. Initially, two algorithms,
Fraud real-world imbalanced datasets were conducting
Detection” selected for experimentation. robustness analyses
Subsequently, the researchers across diverse
optimized the XGBoost model's datasets,
parameters using the investigating feature
 SOFTWARE REQUIREMENTS

• Front End – Anaconda IDE

• Backend – SQL
• Language – Python 3.8
 HARDWARE REQUIREMENTS

• •Hard Disk: Greater than 500 GB

• •RAM: Greater than 4 GB
• •Processor: I3 and Above
 REFERENCES

• 1.Mahesh A. & Ganesh Bhat. (2021). India Digital Payment Service in

India - A Case Study of Unified Payment Interface
• 2. Dinesh M. Kolte, & Dr. Veena R. Humbe (2019). Study of UPI/BHIM
Payment System in India, International Journal of Science and
Research (IJSR)
• 3.Shivam Tripathi & Pranshu Dixit (2020). A Study on Adoption of
Digital Payment through Mobile Payment Application with Reference to
Gujarat State. International Journal of Scientific & Technology Research
Vol. 8, (8)
• 4.Thirupathi M., Vinayagamoorthi D.G.& Mathiraj S.P. (2019). Effect of
Cashless Payment Methods: A Case Study Perspective Analysis
International Journal of Scientific & Technology Research, Vol. 8, (8)
• 5. Sujith T S, Julie C. D. (2017). Opportunities and Challenges of E-
Payment System in India, McGraw, Hill Delhi.