Classifying individual credit scores with LightGBM
Our second case study is a problem of credit score classification for individuals. The dataset is available from https://www.kaggle.com/datasets/parisrohan/credit-score-classification?datasetId=2289007.
The dataset is significantly larger than the previous problem and has unique data formatting problems. For brevity, we will not go through the solution in as much detail as with the previous problem (as much of the work is the same), but the end-to-end solution is available at https://github.com/PacktPublishing/Practical-Machine-Learning-with-LightGBM-and-Python/tree/main/chapter-6/credit-score-classification.ipynb.
Problem definition
The dataset consists of 100,000 rows and 27 columns representing individuals’ demographic and financial information, including a credit score rating. The data includes information regarding individual income, number of loans, payment behavior, and investments. The credit score may be rated...
