Techniques

Numerical range:

Grouping:

Dimenisonality reduciton Feature crossing

Shapley values??

Regardless of the model, feature engineering plays a critical role in the performance of your prediction. Here are some features you might consider:

Historical Prices: Lagged values of past prices.
Rolling Statistics: Rolling means, standard deviations, and other statistics over different time windows.
External Factors: Economic indicators, interest rates, and other macroeconomic variables.
Property Features: Size, number of rooms, location-specific features.
Temporal Features: Time of year, day of the week, seasonality factors.

Some new features I created that inform about the legal state of the

okupas - is the property illegally occupied - read more about the problem in Spain here
alquilado - does it currently have a tenant living inside and the buyer will have to respect the contract until it ends on the date as outlined in the original contract - read more here
propiedad_nuda - when we buy the property but current owners have the right to live there for the rest of their lives. We sometimes need to pay them monthly allowance as well - read more here