A proper procedure that is created for being familiar with machine learning methodology using a dataset related customer credit estimation.
As a early project which has contributed to my understanding of some traps and tricks in machine learning, the steps are relatively simple, including data processing, training and evaluation.
During that period, I didn't figure out the concept of data leakage and different solvers, so some operations are not attempted, like by splitting test-train before any data processing (except cleaning), and a function that compares solvers(“newton-cg”, “lbfgs”, and “sag”).
In this case, only Logistic Regression is implemented now.