ABSTRAT:
Nowadays digitalization gaining popularity because of seamless, easy and convenience use of e- commerce. It became very rampant and easy mode of payment. People choose online payment and e-shopping; because of time convenience, transport convenience, etc. As the result of huge amount of e-commerce use, there is a vast increment in credit card fraud also. Fraudsters try to misuse the card and transparency of online payments. Thus to overcome with the fraudsters activity become very essential. The main aim is to secure credit card transactions; so people can use e-banking safely and easily. To detecting the credit card fraud there are various techniques which are based on Deep learning, Logistic Regression, Naïve Bayesian, Support Vector Machine (SVM), Neural Network, Artificial Immune System, K Nearest Neighbor, Data Mining, Decision Tree, Fuzzy logic based System, Genetic Algorithm etc
EXISTING SYSTEM :
In Card not Present (CNP) fraud, fraudster attempt to mislead the system by dissembling to be some other person. Mail and the web are major routes for fraud against merchants who sell and ship merchandise, and affects legitimate mail order and web merchants. In Skimming, they are obtaining personal data regarding someone else’s credit card utilized in an otherwise normal transaction. There is a tiny device (skimmer) which is used to swipe and store huge amount of victim’s information. In phishing, Scammers might use a range of schemes to lure users into giving them their card info through tricks corresponding to websites simulation to be of a bank or payment system. When card is steal or lost, there are chances for a thief that he make unauthorized transaction before cardholder block the card. The remainder of the paper structured as shown below: Theoretical background explained in section-2. In which various data mining techniques are briefly explain. After that, existing techniques explained in section-3. It includes artificial immune system, Bayesian belief network, logistic regression, decision tree, self-organizing map, hybrid methods, etc. In section-4, we presented analysis of classification techniques with their methodology and challenges. In last section-5 and 6 motivation and conclusion of the paper presented
EXISTING SYSTEM DISADVANTAGES:
1.LESS ACCURACY
2. LOW EFFICIENCY
PROPOSED SYSTEM :
Credit card detection is a fascinating domain. From this survey, we analysed machine learning is best in compare to prediction, clustering, outlier detection etc., that earlier used. Machine-learning techniques are mostly preferred in fraud detection, because of its high accuracy and detection rate. Still researchers are struggling to get more accuracy and detection rate. Moreover, organizations are interested in finding methods that can reduce cost and increase the profit; they can find and select the method from above studies
PROPOSED SYSTEM ADVANTAGES:
1.HIGH ACCURACY
2.HIGH EFFICIENCY
SYSTEM REQUIREMENTS
SOFTWARE REQUIREMENTS:
• Programming Language : Python
• Font End Technologies : TKInter/Web(HTML,CSS,JS)
• IDE : Jupyter/Spyder/VS Code
• Operating System : Windows 08/10
HARDWARE REQUIREMENTS:
Processor : Core I3
RAM Capacity : 2 GB
Hard Disk : 250 GB
Monitor : 15″ Color
Mouse : 2 or 3 Button Mouse
Key Board : Windows 08/10
