In the financial sector, credit lines are the main source of income. The main income source is investing in the amount and collecting interest using the principal amount. However, they are not able to collect all investments because defaulters are not ready to pay the amount. Loan prediction played a vital role in this scenario. By predicting loan defaulters, the institution can reduce the number of faulty account holders. The most popular method is the scientific method, where Machine Learning is used. Machine Learning has statistical models that can perform specific task to predict the credit data of upcoming future. Prediction can be performed using supervised learning techniques such as decision tree, random forests, Gaussian Naïve Bayes, AdaBoost, Support vector machines, and logistic regression. This study aims to build credit scoring by adopting five of them for further study. The primary data are collected from a cooperative financial institution where 13600 data points on credit was collected, 70% of the data was trained, and 30% was separated for testing for test 1. Additionally, same data set is trained and test in 8:2 ratio for test 2.