Anomaly-Detection is the identification of rare items, events or observations which raise suspicions by differing significantly from the majority of the data.
- The use of the latest technology by criminals is becoming increasingly complex with the methods of credit card fraud.
- Unless the customer reports it, it is no longer a simple task to identify outliers/anamolous transactions beforehand.
- This means that when dealing with financial information or sensitive customer data, the use of machine-learning algorithms becomes imperative because these algorithms can examine data volumes including text, images and videos, analyze a pattern, and detect an anomaly in seconds, with high precision.
- Now we have many supervised and uncontrolled methods of learning, but it's a great task to choose the algorithm.
Link to data: https://www.kaggle.com/datasets/mlg-ulb/creditcardfraud