Dark Mode
Financial Fraud Detection Sample
Finance & Banking Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
This collection of data pertains to online credit card activities, specifically focusing on identifying and classifying transactions as either fraudulent or legitimate. The rising volume of online transactions and sophisticated attacks necessitates acquiring deep knowledge and discovering the underlying patterns associated with fraudulent activity. Understanding these patterns is crucial for credit card companies to preemptively stop suspicious transactions and enhance customer security.
Columns
This file contains 20 distinct data fields:
- DOMAIN: The masked domain name corresponding to the customer's email address used during the transaction.
- STATE: The state code representing the customer's location.
- ZIPCODE: The zip code for the customer's location.
- TIME1: The first anonymised hour feature of the transaction.
- TIME2: The second anonymised hour feature of the transaction.
- VIS1: An anonymised feature (#1) associated with the feature set VIS.
- VIS2: An anonymised feature (#2) associated with the feature set VIS.
- XRN1: An anonymised feature (#1) associated with the feature set XRN.
- XRN2: Anonymised feature (#2) for feature XRN.
- XRN3: Anonymised feature (#3) for feature XRN.
- XRN4: Anonymised feature (#4) for feature XRN.
- XRN5: Anonymised feature (#5) for feature XRN.
- VAR1: An anonymised feature (#1) for the feature set VAR.
- VAR2: Anonymised feature (#2) for feature VAR.
- VAR3: Anonymised feature (#3) for feature VAR.
- VAR4: Anonymised feature (#4) for feature VAR.
- VAR5: Anonymised feature (#5) for feature VAR.
- TRN_AMT: The monetary amount of the specific transaction.
- TOTAL_TRN_AMT: The total monetary amount of the transaction.
- TRN_TYPE: The designated type of transaction, marked as either FRAUD or LEGIT.
Distribution
The dataset, typically structured as a CSV file for ease of processing, details 94,682 individual transactions. It is a highly structured, tabular file containing 20 columns. The data is generally complete, with no mismatched or missing values reported across the fields, ensuring high data integrity.
Usage
This data is ideally suited for developing and testing machine learning models focused on Binary Classification. Primary applications include predictive modelling for credit card fraud detection, pattern recognition of suspicious activities, and generating insights into customer behaviour that might expose security vulnerabilities. It can be used to train systems designed to halt fraudulent transactions in real-time.
Coverage
The data includes geographical elements via state codes and zip codes, with some sources indicating a focus relevant to the Russian geographical context. Temporal scope is provided through two distinct hour-based features (TIME1 and TIME2). The records capture activity leading to transactions definitively marked as either FRAUD or LEGIT.
License
CC0: Public Domain
Who Can Use It
- Data Scientists: For building, training, and evaluating predictive fraud models.
- Machine Learning Engineers: For optimising classification algorithms and deploying solutions.
- Financial Risk Analysts: For identifying specific variables that increase fraud exposure.
- Security Researchers: For studying modern cybercrime transaction methods.
Dataset Name Suggestions
- Online Credit Card Fraud Transactions
- Binary Transaction Fraud Data
- Financial Fraud Detection Sample
- CC_FRAUD Tabular Analysis
Attributes
Original Data Source:Financial Fraud Detection Sample
QUALITYLoading...