CREDIT CARD FRAUD DETECTION USING MACHINE LEARNING

PROJECT PHASE I
REVIEW

TEAM MEMBERS
UNDER THE GUIDANCE OF
AGASH S (21TD0103)
Mrs. K. Maheswari Mam
GUNALAN A (21TD0120)
(Professor of the Department /CSE)
LAKSHMIBALAN S (21TD0138)
RANJITH M (21TD0156) 1
INTRODUCTION TO MACHINE LEARNING:
Machine Learning:
• Machine learning (ML) is a branch of artificial intelligence(AI) and computer science that
focuses on the using data and algorithms to enable AI to imitate the way that humans learn,
gradually improving its accuracy.
• A machine learns patterns and relationships from data and uses that knowledge to improve its performance on tasks.
Machine learning methods:
• Supervised learning, also known as supervised machine learning, is deuse of labeled datasets to train
algorithms to classify data or predict outcomes accurately.
• This occurs as part of the cross validation process to ensure that the model avoids over
fitting or under fitting.
• Unsupervised learning, also known as unsupervised machine learning, uses machine learning
algorithms to analyse and cluster unlabelled datasets (subsets called clusters).
• These algorithms discover hidden patterns or data groupings without the need for human
intervention. 2
• Semi-supervised learning offers a happy medium between supervised and unsupervised learning.
• During training, it uses a smaller labeled data set to guide classification and feature extraction
from a larger, unlabeled data set.
Common machine learning algorithms:
• Neural networks simulate the way the human brain works, with a huge number of linked
processing nodes used in image recognition, speech recognition, and image creation.
• Linear regression is used to predict numerical values, based on a linear relationship between
different values. For example, the technique could be used to predict house prices based on
historical data.
• Logistic regression algorithm makes prediction for categorical response variables, such as
“yes/no” answers to questions. It can be used for applications such as classifying spam and
quality control on a production line.
• Clustering algorithms can identify patterns in data so that it can be grouped. Computers can help
data scientists by identifying differences between data items that humans have overlooked.
• Decision trees can be used for both predicting numerical values and classifying data into
categories. Decision trees use a branching sequence of linked decisions that can be represented
3
PROBLEM IDENTIFICATION:
• India’s banking factor has grown rapidly in modern times. Credit card fraud involves
unauthorized. Use of credit card or its data. Credit cards are widely used for online transactions.
• Due to this transaction frequency and records are rising significantly in the digital age, credit
card fraud must be identified to prevent fraud.
• This requires analyzing and verifying client behavior to access find frauds and stay away from
the bad behavior of frauds.
• Credit card fraud occurs when someone else use your card details without your consent, theft of
your credit card, pin or account details allows criminals to make unauthorized purchase without
the card.
• You may have committed the crime to receive the things for free or the illegal money, this
because of lot of money may be made quickly without many dangers, and it will take a long
time to catch the criminals.
• Worlds largest fraud cases involved in e-commerce websites and increasing the risk of credit
4
card fraud, mainly this fraud affects the banks to cover the loss made by the user of the credit
SUSTAINABLE DEVELOPMENT GOAL (SDG)
SDG 8 : Decent Work and Economic Growth
• It emphasizes the importance of equitable economic opportunities, ensuring that all the
people have access to safe, dignified and fair employment while sorting economic
progress that benefits everyone without harming the environment.

Access to Financial Services (Target 8.10)

• Goal : Strengthen the capacity of domestic financial institutions to encourage and expand access to banking,
insurance, and financial services for all.
• Encourages entrepreneurship in FinTech fosters innovation in fraud addiction and contributes to economic growth
through reduce the financial losses.
• Improved security measures and technology: encourage innovation and investment invent tech boost trust, and
increases consumer spending.
• Reduced transaction cost: efficient fraud, deduction, prevention reduce cost of businesses.
• Secure and efficient payment systems: increased consumer spending and confidence, innovation and investment in
financial technology, reduced the cost for businesses.
• E-commerce expansion is completely depending upon the purchasing products by using credit card under let the
economy to be grown.
• By detecting the frauds and credit card can enhance the transaction over the forex while travelling to other
countries. 5
EXISTING WORK:
Basic Operation:
• In my IEEE Base Paper Mrs. Subhashini Peneti initially collected. The data said contains changes made by credit card
holders by Europe nations and then loaded the data said and explored the data.
• Then the process involving data pre processing where they performed three main operations such as formatting,
cleaning and sampling.
• In the formatting method of putting data in a way that most appropriate to apply calculations, cleaning is a very
important procedure at expelling or dealing with lost information and finally sampling is technique which analysis
the subset from large data set, because the large data said cannot provide better results
Data Visualization:

• Future expression is a column reduction process and column selection which makes the significance of the
columns.
• Data Visualization is a technique for graphically or visually presenting data for the analyzed frauds from total
transactions and finally comparing the accuracy of each proposed algorithm and exploiting these algorithms and
mission learning courses from the servers of the bank. 6
PROPOSED WORK:

• Data source: the Kamesh learning repository contains well known data set of credit card from various firms
contains historical data utilized by several researches for their research projects.
• Data pre-processing: in outlined the procedures we need to take to convert or Encore data so that the model could
simply understand it by using the normalizing future engineering and applied sampling technique.
• Data balancing: sampling technique is used for balancing the large set of data and to balance the inbalance dataset
that often arises with the model in classifying data
Machine Learning Techniques:
• By using various machine learning techniques like random forest, logistic regression, Naïve Bayes, K Nearest
neighbors, XG boost producers different accuracy levels.
• By comparing those accuracy levels, we are going to find out the best algorithm for finding the fraud of credit card
with large data set.
• They are also going to propose the neural networks algorithm to compare the accuracy produced
• And also instead of executing these machine learning algorithms from the service of bank, he also propose the
web applications for instant direction of credit card frauds by uploading Excel data sheets of transactions. 7
ABSTRACT:

• Credit card for now used in daily life. The need for online business transaction is rising as quickly as the internet
which influences customers to buy the things they need in to raise enough funds. Can occasionally prevent you from
getting the things.
• There are many types of fraud in our daily life. One of the frauds occurring this days is credit card fraud more than 50
percentage of the Indians are experienced credit and debit card frauds and beyond 30 percentage of individual utilize
these cards.
• That the bank server can track these transaction using machine learning these algorithms detect, and Anpe
transaction fraud.
• This provides a thorough review of effectiveness of various machine learning methods with minimum accuracy of 95
percentage using supervision machine learning techniques in terms of classifying fraudulent and non-fraudulent
transactions with getting the higher accuracy results.
• Each transaction as few functions in the dataset, we shall classify and identify fraudulent transaction from those
functions of dataset.
8
LITERATURE SURVEY:
• New credit card fraud deduction approach is using neuro networks, data mining and distributed data mining. There
are several additional credit card fraud investigation methods, Credit card fraud deduction, using machine learning,
base paper we will improve work into area tools for dealing with uneven data and process for deducting fraud.
• These methods might be classified as sampling. Classification are similar credit card, fraud direction, uses SBM
decision entries, gradient algorithms in 2023 Second international conference for innovation in technology.
• Classification algorithm include Navie bayes algorithm, relies on bayes theorem estimates event, like from available
data.
• The efficient method for recognizing credit card card was put out by E. Esenogho at al and hybrided data resampling
strategy , a neural network and an ensemble algorithm.
• The artificial neural network ANN better across all parameters, training them cause a lot. Every algorithm has such
sustainable drawbacks. This algorithm output changes by environment which is a drawback provided by the 2023
referral paper proposed by Indrani vejjala.
• The survey indicates that normalizing the indolence data set into the balance. Data set is much more difficult by
using ordinary algorithm so that we are using neural network algorithm. 9
REFERENCE
 "Animal Detection using Faster R-CNN for Wildlife Monitoring" by S. Kumar and R. Singh, published
in the Journal of Advanced Research in Dynamical and Control Systems in 2018.
 Kupyn, O., & Prancwhuk, D. (2019). Fast and efficient model for real-time tiger detection in the
wild. In Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops
(pp. 0-0).
 Tan, M., Chao, W., Cheng, J. K., Zhou, M., Ma, Y., Jiang, X., ... & Feng, L. (2022). Animal detection and
classification from camera trap images using different mainstream object detection architectures.
Animals, 12(15), 1976.
 Terven, J., & Cordova-Esparza, D. (2023). A comprehensive review of YOLO: From YOLOv1 to
YOLOv8 and beyond. arXiv preprint arXiv:2304.00501.
 "An Efficient Illumination Invariant Tiger Detection Framework for Wildlife Surveillance" by
Gaurav Pendharkar, A. Ancy Micheal, Jason Misquitta, Ranjeesh Kaippada. arXiv:2311.17552.
 Redmon, J., & Farhadi, A. (2018). Yolov3: An incremental improvement. arXiv preprint
arXiv:1804.02767.
 Liu, B., & Qu, Z. (2023). AF-TigerNet: A lightweight anchor-free network for realtime Amur tiger
(Panthera tigris altaica) detection. Wildlife Letters, 1(1), 32-41.

10
THANK YOU!

11