How it works?
Where does the data come from?
Hesti gathers data from tens of thousands of real estate ads in Latvia over the periods of several months.
How does the algorithm work?
We use gradient boosting models (GBM) with different parameters like: area, room count, construction period, and a generated micro-distance to objects in the location (center, neighborhoods, cities, etc.). As a result, the average market price is shown with a confidence interval of ±10%.
How to interpret the result?
The user is shown the average price of similar properties. The final amount depends on how well the property is maintained and other factors that can influence the price.