Insurance is an integral component of society and must be affordable and reliable to provide everyone with aid in time of need. However, insurance fraud brings disturbance to the insurance sector, increasing the cost for both insurers and the insured. Frauds in Insurance has been around since the inception of Insurance organizations. Different types of insurance are prone to different types of frauds/crimes, however, in most cases, it manifests deliberate damage to the insured person/item for the purpose to obtain claims from the insurance provider. This could be a white lie on an insurance application, an exaggerated claim for damages or injuries, or even a claim that is blatantly false.
         The traditional approach of detecting such frauds is still being used by few insurers which require human interventions. Some of the traditional methods are internal audits, red flag indicators, and scoring models. Detecting insurance fraud and the losses can be difficult since not every claim can be investigated thoroughly. Hence, the portion of fraudulent claims that are detected is smaller than that of the portion of fraudulent acts that are committed.
          The dataset used for model training and evaluation contains only 24.7% fraudulent claims compared to 75.3% of non-fraudulent claims.
           The project has attempted to identify such fraudulent claims with help of historical data. fraud detection is a knowledge-intensive activity that allows classifying correctly whether the claim is legitimate or fraudulent. The project would help to make it easier for human agents to investigate a fraudulent claim.

Dataset

An auto insurance is a policy purchased by vehicle owners to mitigate costs associated with getting into an auto accident. Instead of paying out of pocket for auto accidents, people pay annual premiums to an auto insurance company. the company then pays all or most of the costs associated with an auto accident or other vehicle damage.

Dataset Statistics
Number of variables	39
Number of observations	1000
Missing cells (%)	0.0%
Duplicate rows (%)	0.0%

Link : https://www.kaggle.com/roshansharma/insurance-claim