The project, which was sponsored by Professor Shuhua Wu, provided me with experience using Decision Tree analysis on a dataset. The main challenge was to use a Jupyter Notebook to create a classification model that predicted an outcome variable. I needed to choose an appropriate dataset, explain the features, and then modify the algorithm by determining the best value for the minimum split. This step was essential for avoiding overfitting. I learned how to thoroughly interpret the final tree structure, particularly by analyzing the nodes with the highest and lowest probability, in order to acquire actionable insights into the feature rules that drive the prediction. Successfully identifying sample data points tested my comprehension of how Decision Trees use data to provide specific business rules.
Connect and innovate with an elite information systems program
