Forecasting Italian SME defaults

An Application of Random Forest!

Introduction

• Objective: Project for my intern at Research Center VERA, Ca' Foscari University of Venice.

• Abstract: Developed a random forest model based on a highly unbalanced dataset of 10K Italian SMEs (labeled with default/active businesses) to forecast which is going to default with 87% accuracy

• Status: Completed

Methods Used

• Random Forests (Highly Unbalanced Dataset)

Dependencies

• randomForest
• ggplot2
• plotly