Credit Approval System: SVM Model (Built from scratch and compared against python sklearn fn)
We all have used a credit card for shopping, paying bills or for building a good credit score. It has been an essential part of our modern-day lives. In this report, we aim to predict if an applicant’s credit score is good or bad based on the applicant’s details such as age, work experience, number of family members, housing type etc. The dataset used to train and test the model is extracted from an open source – Kaggle. We found that the model trained can predict if the applicant’s credit is good or bad with an accuracy over 97%.