• Checked Target Incidence using a Binary classifier Model which provides the output in ‘0 - the donor will not give blood’ and ‘1 - the donor will give blood’. Result is [‘0’ - 0.762] & [‘1’ - 0.238]
• Zeroed into LogisticRegression model using TPOT, a Python Automated Machine Learning tool that optimizes machine learning pipelines using genetic programming, giving us the AUC score of 0.7850
• Improved the AUC score by 0.5% to 0.7891, after applying log normalization on Monetary (c.c. blood) feature

Link to GitHub Repository

• Binary Classifier
• Regression