Auto Insurance Claim

The auto-insurance industry is witness paradigm shift. since auto insurance company consist of homogeneous good thereby making it difficult to differentiate product A from product B ,also companies fight for price war,on top of that distribution channel is shifting more from traditional insurance brokers to online purchase which means ability for companies to interact through human touch points is limited and customers should be quoted at a reasonable price .A good price quote make customer to purchase policy and make company to increase profits .Also insurance premium is calculated based on more than 50+ parameters which mans that traditional business analytics based are now limited in their ability to differentiate among customers based on subtle parameter

1.Conquering Market Share
2.Risk Management
3.Increase Profit

The project involve use of dataset with 600k training data and 57 features .In train and test data ,features that are belong to similar group are tagged in feature name ex:(ind,reg,calc,car).In addition feature name include postfix bin to indicate binary feature and cat to categorical feature .feature without these designation are either continous or ordinal

Acknowledgements

EDUREKA