Python Portfolio

Predicting House Prices

This notebook explores the famous Boston housing price dataset and builds a machine learning model to predict average home prices in a neighborhood based on characteristics about that neighborhood. Data visualizations are built with the Matplotlib, Seaborn, and Plotly libraries. The Pandas, Numpy, and Sklearn libraries were used for data preprocessing and the model was built with the Statsmodels library to identify coefficient standard errors. The final model was trained using Ordinary Least Squares regression with a log-linear dependent variable, resulting in an adjusted R-Squared of 0.793. The independent variables that were statistically significant predictors (at the 0.05 level) of average neighborhood price include: crime rate, proximity to the Charles River, pollution level, average number of rooms, commute distance, highway access, average property taxes, average student-teacher ratio, share of minority residents, share of low-income residents.