Do I need to scale my target variable in regression analysis? I have a data which has some features (some kind of products, country in which it is sold etc.) and a Target Variable (the product's price in the country's local currency). So, do I have to scale my Target variable with respect to one common currency for regression analysis?
 A: If you are going to treat a single “price paid” column as the target, you should convert to a single currency prior to producing your model.
For example, an item that costs 1000.00 euro  would likely be very different in “value” compared to an item that costs 1000.00 lek. A model that treats both of these as the same thing numerically will miss the point of the vastly different value of these two things. You could always run your analysis without converting if you were more interested in number paid versus the much more likely value paid.
If, on the other hand, you want to perform your analysis within the bounds of each currency separately, you could avoid converting to  the single, standard currency of your choosing.
In the case that you are most interested in producing a model where currency needs to be converted, I’d recommend considering a curency that your audience will have some familiarity with as the one to convert to. This can help with understanding since they’ll have a frame of reference.
A: You could also add dummy variables that specify the currency in which it was sold. A simple linear model for two currencies (USD and EUR) and two products (TVs and Computers) would look like this:
local price = a1 * TV + a2 * USD + error
where a1 and a2 are constants, TV and USD are dummy variables. TV takes a value of 1 when your product is a TV and 0 otherwise. USD takes a value of 1 when the currency is USD and 0 otherwise. 
You can try standardizing your target variable but interpretation would not be as straightforward.
