I am writing a model that can predict the price of a car with my own dataset. Everything in the data is fine, except the version of car. They are so specific that I am not sure any model can handle that many details (and about 15% of versions are absent). I have from 30 to 10,000 examples of every model of every car, depending on its popularity (the whole dataset is 270,000). Here is a photo to give you some clue. So I wanted to ask what should I do with such diverse data. You can take a look at the whole dataset here. Continue reading...
