The real estate market is a very complex system that is influenced by a huge variety of factors. Being able to predict house prices can be a challenging problem, but it can also be extremely useful (and fun), especially for people like us who live in Milan. In this this project, we aim to use machine learning techniques in order to predict house prices in Italy. We will first conduct some exploratory data analysis to gain insights and solve any potential issues with the given data, and then proceed to add some new features to the model that can be exploited at prediction time. Next, we will train and evaluate different machine learning algorithms and use cross-validation to select the best-performing one. We will finally conclude with a mention to some possible limitations for the analysis, aswell as ideas for future improvement.
- Notebook: Contains all the code. From EDA and approach with missing values and outliers, to feature engineering and machine learning models.
- PDF Report: Four pages summary containing explanations and interpretations for the strategies and procedures adopted.
Please note that the dataset used in this project is private and cannot be made available.