Final Project - Data Science: Barcelona Property Price Prediction

This report details the development of a regression model for predicting property prices in Barcelona, based on a dataset of 413 residential properties. The project aimed to predict prices for an additional 200 properties. Through a reduction method, the initial model, including various independent variables such as floor size, number of rooms, and location, was refined to include only statistically significant variables. The final model, with seven variables, explained 78.6% of the variance in property prices. Key findings revealed that factors like floor size, number of bathrooms, 'Atico' status, parking, yard, and location in Sarria-Sant Gervasi significantly influenced property prices. The model's coefficients and statistical significance were evaluated, and the Root Mean Squared Error (RMSE) was used to assess the accuracy of the price predictions for the additional 200 properties, as per the assignment brief.