Detailed analysis on fraud claims in insurance companies, gives you information as to why huge loss take place in insurance companies

Introduction ( Purpose )

Fraud detection occurs in many industries such as banking and financial sectors , insurance , healthcare and more. Upcoding fraud in recent years has risen sharply where fraudsters come up with different ideas to claim a financial gain through insurance claims . In Upcoding fraud by claiming more amount than the usual costs for their service. Incorporating artificial intelligence with data mining and statistics help decrease these kinds of frauds. Data mining is used to scale huge transactions and detect the fraudulent ones whereas the hybrid learning methodology helps detect frauds.

The primary incentive to commit upcoding is financial gain. Upcoding appears in different ways such as Upcoding of services, Upcoding of items and Duplicate claims. Data mining helps detect such fraudulent claims in the future. It also increases an adjuster’s efficiency by narrowing down prospective audits and Identifies and isolates factors that indicates potential fraudulent activity.

REQUIREMENTS

--> Functional Requirements

1. The model should be able to detect the fraud transactions .

--> Non Functional Requirements

1. The accuracy of the predicted value must be precise.
2. The model should never fail in the middle of operation.
3. The model should work consistently across various platforms.

--> Software Requirements

1. OS Version: Windows 7(64 bit) or newer versions.
2. Coding Language: Python 3.6
3. Platform: Jupyter Notebook

--> Hardware Requirements

1. Processor: i5 or i7 Intel Processor
2. Primary Storage: 8 GB RAM or above (Recommended 16 GB)
3. Secondary Storage: Any standard HDD or SDD