Compares two different sets of modelling, one set select variables with/out any categorization
Consumer Lending Segment is very complex exercise for banks so they often offload to create profiling of different customers to different service provider and use their credit ratings of individuals to disbursement of loans. In this model we have extensive data of different customers from age to marital status to education to employment and so on . This model compares two different sets of modelling, one set select variables with out any categorization and one have categorization, followed by variables whom we think would matter most. It clearly shows the amount and spread of data determines the outcome of the model rather than what we think, what matter most. Suggestion are most welcome to prune it further and make it more useful to community. One more thing, reference of this data is from the internet itself.