Machine learning in practice
Real data is not a tidy table of numbers.
Categories and features
The same flats, turned into numbers in different ways.
The data as the model sees it
The first rows, after the choices above.
Predicted against real price, validation flats
The weights
One weight per column.
Your tries