Fraud-Detection

Detect behavioural anomalies to identify fraudulent users

Performs One-Class Classification

Final Version.ipynb contains all the updated code

Encode labels for categorical data -> Assigns numbers to categories
Extracts more features from dates
Merge fraudsters data with the transactions they have done to derive a pattern for each transaction
Since this is one class classification, OneClassSVM is perfect for this situation, derives a pattern for fraud transactions
Save all the users whose transactions were tagged by the SVM as fraudster

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
.gitignore		.gitignore
Cleaner Clustering.ipynb		Cleaner Clustering.ipynb
Final Version.ipynb		Final Version.ipynb
Psuedo Clustering.ipynb		Psuedo Clustering.ipynb
README.md		README.md
countries.csv		countries.csv
currency_details.csv		currency_details.csv
fraudsters.csv		fraudsters.csv
fx_rates.csv		fx_rates.csv
updated_fraudsters.csv		updated_fraudsters.csv
users.csv		users.csv