You can change the data type of the variable with Var Type Transform in the menu on the Data tab then re-select it. Also, I believe you mentioned previously that you want to bypass the limit on needing at least 20 minority class examples, so upsampling (essentially raw data + simulation or basic resampling) would be a good approach.
I see on your screenshot two possible values with high skewness. May you tell more about the reason why you decided to have this as a regression problem? And what regression metric do you assume for this project?