FinHub creates case studies for students to apply data science in a practical matter. These case studies are based on real world data and students are expected to utilise machine learning to develop predictive analytical tools to solve them.
Automating Credit Card Approvals
This case study requires the student to develop a model that is able to mimic credit card approvals like how a human would, classifying it as approved or rejected.. The goal is to achieve a performance level very similar to humans based on accuracy, precision, and recall.
Iowa Housing Prices
This case study aims at predicting the sale price of Iowa houses based on factors about the house, which is a regression problem. These factors include number of stories, size of the house in square feet, neighbourhood, and others.
This case studies aims to classify countries based on how "risky" they are. This classification is done by looking at factors of 122 countries. These factors include corruption, peace, legal, and GDP growth rate.