Automated valuation models

Abstract

This work is an introduction to AVM (automated valuation models), the different models to predict automatically the value of a property (and give a bound of the error). This work aims to show the different options of models and their advantages and disadvantages and also to show consistently the math behind them with the final objective of being able to know which model is the best in different situations. Also, we will study some techniques used nowadays by appraisers and see if we can improve them. We will also prove our theory on some examples with real data provided by Catsa.