Two of our consultants organized a predictive analytics ‘hackathon’ for the the Executive Programme in Business Analytics at Solvay Business School. During this event, the students fitted machine learning algorithms to find the best subset of potential donors to target in an NGO’s direct mail campaign. Students learned about different algorithms such as decision trees and logistic regression, about the difference between training, test, and validation set, and about different ways to evaluate predictive performance such as the confusion matrix and the lift curve. At the end of the two-day hackathon, they presented their results to the program director acting as the CEO of the NGO. Impressed by the results, she decided on the spot that she would hire all of the business analysts in the class!
