Skip to content
#

tp-fp

Here are 2 public repositories matching this topic...

Language: All
Filter by language

Credit risk is an inherently unbalanced classification problem, as the number of good loans easily outnumber the number of risky loans. I employed Machine Learning techniques to train and evaluate models with unbalanced classes. I used imbalanced-learn and scikit-learn libraries to build and evaluate models using resampling. I also evaluated the…

  • Updated Nov 19, 2020
  • Jupyter Notebook

Add this topic to your repo

To associate your repository with the tp-fp topic, visit your repo's landing page and select "manage topics."

Learn more