Credit card fraud detection using

machine learning
Abstract
Reality-based credit card fraud detection is the
project's primary emphasis. There has been a
dramatic increase in fraudulent acts as a result of
the meteoric rise in the use of credit cards in recent
years. The goal is to gain services or items without
paying for them, or to steal money from an account
without permission. All credit card issuing
institutions must now have reliable fraud detection
systems to cut down on losses. The fact that neither
the card nor the cardholder need to be present
during the transaction poses a significant obstacle
to the success of the firm.
Since the shop can't check the customer's identity, it's
feasible that a fake card may be used. The suggested
method employs Machine Learning techniques to
increase the precision with which fraud is detected.
The random forest method is used for classification,
analyzing both the input dataset and the user's most
recent input dataset. In the end, you should maximize
the precision of the obtained information. Precision,
sensitivity, and accuracy are the metrics used to assess
the effectiveness of the methods. The graphical model
is then seen after some of the given characteristics are
processed to identify fraud. Accuracy, sensitivity,
specificity, and precision are used to assess the
methods' effectiveness.
INTRODUCTION
 The migration of business to the Internet and the electronic
monetary transactions that occur in the continuously
growing cash-less economy have made the accurate
detection of fraud a significant factor in securing such
transactions. Credit card fraud occurs when a thief uses
credit card information to complete purchase processes
without permission from the credit card owner. The large-
scale use of credit cards and the lack of effective security
systems result in billion-dollar losses to credit card fraud.
Because credit card firms are typically unwilling to
announce such facts, it is difficult to obtain a precise
approximation of the losses.
EXISTING SYSTEM
 In existing System, a research about a case study involving
credit card fraud detection, where data normalization is applied
before Cluster Analysis and with results obtained from the use of
Cluster Analysis and Random forest on fraud detection has
shown that by clustering attributes inputs can be minimized.
And promising results can be obtained by using normalized data
and data should be Random forest trained. This research was
based on unsupervised learning. Significance of this paper was
to find new methods for fraud detection and to increase the
accuracy of results.
Disadvantage:
 1. In this paper a new collative comparison measure that
reasonably represents the gains and losses due to fraud
detection is proposed.
 2. There is a need for high accuracy models
PROPOSED SYSTEM
 In proposed System, we are applying Support vector
machine algorithm for classification of the credit card
dataset. SVM is an algorithm for classification and
regression SVM has advantage as it corrects the habit of
over fitting to their training set.
 Advantages of proposed system:
• The results obtained by the SVM is best compared
to any other Algorithms.
• SVM give high accuracy
REQUIRMENTS
SOFTWARE REQUIREMENTS
 The functional requirements or the overall description
documents include the product perspective and features,
operating system and operating environment, graphics
requirements, design constraints and user documentation.
 The appropriation of requirements and implementation
constraints gives the general overview of the project in
regards to what the areas of strength and deficit are and
how to tackle them.

 1)Anaconda 3.7
 2)Jupyter Notebook
 3)Python IDLE ( Python 3.7 )
REQUIRMENTS
HARDWARE REQUIREMENTS
 Minimum hardware requirements are very dependent on
the particular software being developed by a
given Enthought Python / Canopy / VS Code user.
Applications that need to store large arrays/objects in
memory will require more RAM, whereas applications that
need to perform numerous calculations or tasks more
quickly will require a faster processor.

 1)Operating System : Windows Only

 2)Processor : i5 and above
 3)Ram : 4gb and above
 4)Hard Disk : 50 GB
SYSTEM ARCHITECTURE
ALGORITHMS
 SVM Algorithm: “Support Vector Machine” (SVM) is a
supervised machine learning algorithm that can be used for
both classification or regression challenges. However, it is
mostly used in classification problems. SVMs are used in
applications like handwriting recognition, intrusion
detection, face detection, email classification, gene
classification, and in web pages. This is one of the reasons
we use SVMs in machine learning. It can handle both
classification and regression on linear and non-linear data.
DATAFLOW DIAGRAM
UML DIAGRAMS
USECASE DIAGRAM Upload credit card dataset

Generate train & test model

Run Random forest algorithm

user
Run proposed SVM algorithm

Detect fraud from test data

Clean & fraud transaction detection

graph
CLASS DIAGRAM
Upload credit card dataset
Generate Train & test model Run Random forest algorithm
dataset
generate model() run()
upload()

Run proposee SVM model

run()

Detect fraud from test data

dataset

detect fraud()

Clean & fraud transaction detection graph

ACTIVITY DIAGRAM
SEQUENCE DIAGRAM
User System

Upload credit card dataset

Generate train & test model

Run random forest algorothm

Run proposed SVM algorithm

Detect fraud from test data

Clean & fraud transaction detection graph

COLLABORATION DIAGRAM
User

1: Upload credit card dataset

2: Generate train & test model
3: Run random forest algorothm
4: Run proposed SVM algorithm
5: Detect fraud from test data
6: Clean & fraud transaction detection graph

System
COMPONENT DIAGRAM
Upload credit Generate train Run Random
card dataset & test model forest algorithm

Run SVM
algorithm

Detect fraud
from test data

Check & fraud transactio

n detection graph
DEPLOYMENT DIAGRAM

User System
CONCLUSION
 The SVM algorithm will perform better with a larger
number of training data, but speed during testing and
application will suffer. Application of more pre-processing
techniques would also help. The other algorithms still
suffers from the imbalanced dataset problem and requires
more preprocessing to give better results at the results
shown by other algorithms is great but it could have been
better if more preprocessing have been done on the data.
FUTURE SCOPE
 In the future ,we will work to improve the model,
considering the features of more dimensions besides
frequency, and try to apply the model on image data.
REFERENCES
 [1] X. Zhang, Y. Han, W. Xu, and Q. Wang, ``HOBA: A novel
feature engineering methodology for credit card fraud detection
with a deep learning architecture,'' Inf. Sci., May 2019. Accessed:
Jan. 8, 2019.

 [2] N. Carneiro, G. Figueira, and M. Costa, ``A data mining based

system for credit-card fraud detection in e-tail,'' Decis. Support
Syst., vol. 95, pp. 91101, Mar. 2017.

 [3] B. Lebichot, Y.-A. Le Borgne, L. He-Guelton, F. Oblé, and G.

Bontempi, ``Deep-learning domain adaptation techniques for
credit cards fraud detection,'' in Proc. INNS Big Data Deep Learn.
Conference, Genoa, Italy, 2019, pp. 7888.