This Solution is built using SQlite3 which is integrated in Python. The information received from the bank, has been split into different tables of SQLite. The idea is to have the database updated in real time and extraction of any fraudulent data which is identified based on certain guidelines in this case which is assumed to be three of them.
- The first anomaly is identified when there is any discrepancy for amount withdrawals less than a certain amount from the user’s account which is unlikely to be have taken place.
- Second Anomaly is based on Frequency. If there are transaction on the same date multiple times.
- Third Anomaly is based on the Preferred location of an individual user which is gathered as per the trend provided.
Check out for above Case study in below mentioned weblink.