Hello, I need to make a project to find Anomaly Detection in Bank Transactions using Unsupervised Learning (Clustering). I do not know if you are familier about Anomaly Detection. Such as
- high transaction,
-more than 5 transaction,
-transaction from one country, at the same time transaction from some other country
-other useful features...
The columns could be : transaction_id | account id | transaction_date | transaction_type | transaction_amount |
recipient_id | recipient_country | client_ip_address | client_card_mask | recipient e-mail
client_card_country | balance | balance after trans |
[login to view URL]
Most of the people using above sample for Anomaly Detection. But I need more professional job. if you wanna do the job. Can you please explain me. What is your strategies. What Anomaly detection can you find?
11 freelancers are bidding on average $29 for this job
Hi, I am Ibrahim, and I am a data scientist, I can help you with anamoly detection, there are several methods to detect, as the data is not labelled, I would run clustering algos. Regards, Ibrahim Anjum
I have been working as a AI Developer for more than 3 years and have done one of the project similar to this where I had used Mahalanobis Distance to detect the anolmaly.