This repository includes the projects related to creating an API off of US Census Data, and using machine learning tools to predict delinquincy of the people who have borrowed loans.
This repository includes some assignments and labs that I completed as part of a course called "Machine Learning for Public Policy" (Spring, 2017). Some of the tasks that I completed and the skills that I learned are:
using the U.S. Census Bureau API to scrap data regarding socio-economic characteristics of different neighborhoods (tracts/blocks) in Chicago. (PA1)
pre-processing data (dealing with missing values, one-hot encoding, standardization and selecting important features) to make it ready for the machine learning phase. (PA2)
using different machine learning techniques (logistic regression, SVM, random forest, gradient boosting) to predict delinquincy of loan borrowers. (PA3/Labs)
applying evaluation metrics (AUC, ROC, and Precision/Recall rates) to evaluate the accuracy and predictive power of the machine learning model. (PA4/Labs)