Project Last Mile: Analytics Case Challenge

About the Competition Project LastMile is an analytics and machine learning challenge designed to test participants' ability to analyse data, identify key drivers, develop predictive models, and derive actionable business insights. Participants will work with a provided dataset to build prediction models and compete based on their model performance on a hidden evaluation dataset. Competition Structure Round 1: Online Prediction Challenge Participants will develop predictive models using the provided training dataset. Predictions must be submitted in the prescribed CSV format. Evaluation metric: ROC-AUC. The competition will be hosted on Kaggle. Top 8 teams will qualify for the next round. Round 2: Online Presentation Qualified teams will present their analysis and recommendations. Presentation: 8 slides. Pitch: 10 minutes. Q&A: 5 minutes. Eligibility First-year and second-year students. Cross-year teams are permitted. Team size: 2–3 members. Important Instructions Participants must submit original work. Qualified teams must submit their code/workbook within 24 hours of notification. Teams must be able to reproduce their submitted results. Failure to reproduce results may lead to disqualification. Further details, datasets, submission instructions and deadlines will be communicated to registered participants.