Skip to content

A banking customer wanted to provide self-service facility for end-customers to predict success of their loan applications. The given data had to be cleansed and a prediction model had to be built. Data was pre-processed by removing redundant and zero-impact variables and applying label encoder to encode non-numeric variables. Normalization was …

Notifications You must be signed in to change notification settings

vickytr44/Loan-approval-prection

Repository files navigation

Loan-approval-prection

Loan approval prediction: hackathon by analytics vidhya

Steps to create API

  1. Run the model.py file in terminal (make sure train data filled.csv path is correctly provided in model.py) python model.py
  2. If model.py is run successfully then model.plk file will be created in the directory
  3. Run server.py and provide port number along with it python server.py 12345
  4. If server .py is run successfully then API is set up in http://127.0.0.1:12345/predict

To test API, Please provide the following json data input: [ {"Credit_History": 1, "Total_Income": "100000", "LoanAmount": "10000"}, {"Credit_History": 1, "Total_Income": "5000", "LoanAmount": "100000000"}, {"Credit_History": 0, "Total_Income": "2000", "LoanAmount": "100000"}, {"Credit_History": 0, "Total_Income": "10000000", "LoanAmount": "100"} ]

output: { "prediction": "['Y', 'N', 'N', 'Y']" }

About

A banking customer wanted to provide self-service facility for end-customers to predict success of their loan applications. The given data had to be cleansed and a prediction model had to be built. Data was pre-processed by removing redundant and zero-impact variables and applying label encoder to encode non-numeric variables. Normalization was …

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages