The objective of this notebook is to build the best model to predict if a company is going to default using deep learning models. We will explore hyper parameter tuning, down-sampling and over-sampling.
Download the data from https://www.kaggle.com/datasets/fedesoriano/company-bankruptcy-prediction.