Housing is a composide good, because it has many characteristics and those characteristics cannot be sold separately.
To estimate a house price, we first need to identify the characteristics, such as the location, the neighborhood, environmental issues, the house's physical characteristics. (i.e. the structure and interior, the size in square meters, if it has a garden, a garage, etc etc.)
Many of the characteristics of housing are non-market goods, this means there's no market price associated or related with them.
Serra da Estrela - Portugal
To estimate those prices Hedonic regression models, are very useful, as they can isolate non-market goods from the remaining characteristics, and study them separately.
Hedonic models use regression analysis to break the good in question into different parts and determine the value of each part.
The total selling price of a house is reflective of the households willing to pay for this collection of goods.