A project based on Insurance data, to create the best model to predict if a client would earn more or less than $50,000, based on different attributes. Project was for MSC Applied Data Science - Scripting module.
Objective of the project was to compare 5 different models, to establish which is the best one. Upon identifying best model, an ensemble model was created to see if the accuracy could be improved.
Models compared were Decision Tree, Random Forest, Logistic Regression, K-Nearest Neighbour and Support Vector Machines.
Resulting grade: 82.5% (Distinction)