Scribd
Upload
33 views17 pages

Final Project Phase-I Presentation

The document presents a project on UPI Fraud Detection using Machine Learning, detailing the problem of increasing fraudulent transactions in India's UPI system and the limitations of existing detection methods. It proposes a new system utilizing Convolutional Neural Networks (CNN) to enhance accuracy and adapt to evolving fraud patterns. The presentation outlines the project's objectives, system architecture, functional modeling, and concludes with a clear plan for the next development phases.

Uploaded by

Prerana Patil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
33 views17 pages

Final Project Phase-I Presentation

The document presents a project on UPI Fraud Detection using Machine Learning, detailing the problem of increasing fraudulent transactions in India's UPI system and the limitations of existing detection methods. It proposes a new system utilizing Convolutional Neural Networks (CNN) to enhance accuracy and adapt to evolving fraud patterns. The presentation outlines the project's objectives, system architecture, functional modeling, and concludes with a clear plan for the next development phases.

Uploaded by

Prerana Patil
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 17

P. S. G. V. P.

Mandal’s

D. N. Patel College of Engineering, Shahada


Department of Computer Engineering

A Project Phase-I Presentation on

“UPI Fraud Detection Using ML”


Under the Guidance of Submitted by
Prof. A. B. Koli Ms. Prerana Patil
Ms. Bhakti Patil
Ms. Nikita Patil
Agenda
 Problem Definition

 Existing System(s)

 Proposed System

 Functional Modelling

 UML Modelling

 Conclusion

 References
Problem Definition
 UPI, a mobile-based payment system, has become the backbone of digital
transactions in India. while offering convenience and accessibility, the surge
in UPI transactions has exposed vulnerabilities to fraudsters. the problem
lies in identifying and preventing fraudulent UPI transactions before they
impact users and the financial system.
 The rapid adoption of UPI has led to a parallel increase in fraudulent
activities, posing a significant threat to user trust and financial stability.
 Fraudulent transactions often involve sophisticated techniques like account
takeovers, phishing scams, and unauthorized payments, making detection
challenging.
 The need for real-time fraud detection is crucial to prevent financial losses
and ensure the integrity of UPI transactions.
Existing System(s)
 Rule-Based Techniques: Use fixed rules like transaction limits, location
changes, or frequent transactions to detect fraud.
 Behavioural Analysis: Monitors user habits, like spending patterns, login
times, or device usage, to identify unusual activity.
 Network Analysis: Detects patterns in account-to-account transactions
to find fraud networks or coordinated activity.
 Multi-Factor Authentication (MFA): Adds security layers like OTPs or
biometric checks to ensure user identity during transactions.
Existing System(s)
 Limitations:-
While current UPI fraud detection systems are effective to a degree,
they suffer from challenges like false positives/negatives, reliance on
historical data, adaptability issues, high resource requirements, and user
compliance gaps. These limitations highlight the need for continuous
improvement, innovation, and integration of advanced technologies like AI.
Proposed System
Objectives.
 To develop a system for automatically detect fraudulent transactions in
UPI using CNN.
 To improve accuracy and speed of fraud detection compared to traditional
methods.
 To build a system that can adapt to new patterns of fraud over time.
 To minimize false positives and ensure legitimate transactions are not
interrupted.
Proposed System
System Architecture
Proposed System
System Modules
Registration.
Login.
Dataset Collection.
Preprocessing.
Feature Extraction.
Classification.
Result.
Proposed System
Applicability
Banking and Financial Sector: Enhances the ability of banks to manage
risks associated with digital payments.
UPI fraud application: Helps platforms like Google Pay, PhonePe, and
Paytm detect anomalies.
Cybersecurity Applications: Enhances fraud detection systems by using
CNNs to analyze transaction patterns and classify them as fraudulent or
legitimate.
Functional Modelling
 Data Flow Diagram: Level 0 .
Functional Modelling
 Data Flow Diagram: Level 1 .
UML Modelling
 Use Case Diagram.
UML Modelling
 Activity Diagram.
UML Modelling
 Class Diagram.
Conclusion
 In this phase, we focused on creating documents and plans for the
project. We studied the problem in detail, identified possible types of
fraud, and looked at existing methods to detect them. We also defined
the project’s goals, what it will cover, and the main tasks to ensure
everything is practical and achievable. This phase has created a clear
plan and a strong understanding of the project, making it ready to move
into the next steps of building and testing the system.
References
Aleskerov e, freisleben, b., and, rao b (1997) cardwatch: a neural network-
based database mining system for credit card fraud detection. in conference
(pp. 220–226). ieee, piscataway, nj
Abdallah a, maarof ma, zainal a (2016) fraud detection system: a survey. j
netw comput appl 68:90–113
Artís m, ayuso m, guillén m (1999) modeling different types of automobile
insurance fraud behavior in the spanish market. insurance math econ 24:67–
81
Bolton rj, hand dj (2001) unsupervised profiling methods for fraud
detection. in conference on credit scoring and credit control 7, edinburgh, uk,
5–7 sept
Dorronsoro jr, ginel f, sanchez c, santa cruz c (1997) neural fraud detection
in credit card operations. ieee 8:827–834
Thank You!!!

You might also like