EXTREME GRADIENT BOOSTING TECHNIQUE FOR
ONLINE
PAYMENT TRANSACTION FRAUD DETECTION
ABSTRA
CT :
In today's world, people depend on online transactions for almost everything. Online
transactions have their own merits like easy to use, feasibility, faster payments etc., but these
kinds of transactions also have some demerits like fraud transactions, phishing, data loss, etc.
With the prevalence of mobile e-commerce, fraudulent transactions conducted by robots are
becoming increasingly common in mobile payments, due to which there is a constant threat for
frauds and misleading transactions which can breach an individual's privacy.
It has become a difficult problem for mobile applications to identify robotic automation
accurately and efficiently from a massive number of transactions.
Hence, we presented a machine learning - based transaction fraud detection model with some feature
engineering. The algorithm can get experience; improve its stability and performance by processing
as much as data possible.
In this, the dataset of certain transactions which is done online is taken. Then with the help of
machine learning algorithms, we can find the unique data pattern or uncommon data patterns which
will be useful to detect any fraud transactions. For the best results, the XGBoost algorithm will be
used which is a cluster of decision trees.
The knowledge is applied to extract risk features from transaction data, risk features together with
normal features are input into the boosting algorithm to perform training, and therefore we
incorporate boosting algorithm with prior human knowledge to improve the performance of the
model.
INTRODUCTI
ON :
In today’s world, we are on the verge to become a cashless world. According to various surveys and
researches, people performing online transactions has increased a lot, it’s expected that in future years this
will go on increasing.
Now, while this might be exciting news, on the other-side fraudulent transactions are on the rise as well.
Even due to various security systems being implemented, we still have a very high amount of money being
lost due to fraudulent transactions.
Online Fraud Transaction can be defined as a case where a person uses someone else’s credit card for
personal reasons while the owner and the card-issuing authorities are unaware of the fact that the card
is being used.
“Fraud detection involves monitoring the activities of populations of users to estimate, perceive or avoid
objectionable behavior, which consists of fraud & intrusion. “
This is a very relevant problem that demands the attention of communities such as machine learning and data science
where the solution to this problem can be automated.
In real world examples, the massive stream of payment requests is quickly scanned by automatic tools that determine
which transactions to authorize. Machine learning algorithms are employed to analyze all the authorized transactions
and report the suspicious ones.
These reports are investigated by professionals who contact the cardholders to confirm if the transaction was genuine
or fraudulent. The investigators provide feedback to the automated system which is used to train and update the
algorithm to eventually improve the fraud-detection performance over time.
ADVANTAGE
S:
In this project, we have proposed a model which can predict whether the transactions are fraud or not. We have used
the XGBoost algorithm which has a lot of advantages over other algorithms. Let us understand better—
XGBOOST ALGORITHM: Extreme Gradient Boosting (XGBoost) is an open-source library that provides an
efficient and effective implementation of the gradient boosting algorithm. This algo. is based on decision trees. With
the help of these decision trees, it classifies whether the data is fraud or not.
As it is based on a decision tree , it can give us pretty good accuracy and also efficiency. XGBoost (Extreme gradient
boosting) is one of the well-known gradient boosting techniques having enhanced performance and speed in tree-
based (sequential decision trees) machine learning algorithms, which provides better results and imposes security.
As a result of implementation of this purpose, the occurrence of high-velocity attacks from bots, malicious log-in
attempts and much more is gradually stopped on a larger scale.
SCOP
E:
The scope of this application is very far reaching. Using this system,
the algorithms will be built to go through the dataset and provide the
appropriate output.
In the long run, this system will be quite beneficial as it provides an
efficient system to create a secure transaction system to analyze and
detect fraudulent transactions.
The gradient boosting algorithm attempts to accurately predict a
target variable by combining the estimates of a set of simpler,
weaker models. This accuracy can be increased further by providing
a huge dataset for model training
This system can be used to detect the features of fraud transactions
in a dataset which is very well applicable in various sectors like
banking, insurance, e-commerce, money transfer, bill payments, etc.
This will indeed help to increase security.
THANK
YOU!
