Hi Fellows,
Let say I want to model a country’s house prices for example Australia.
How many models should I use? One model for the whole country, or a model for each state or even one for each district? How should i approach this problem?
I would recommend Tom’s suggestion. Your questions are very reasonable but there is no information about the data to discuss any further. If you have a dataset, why don’t you just try modeling it, or even analyze the dataset and try to find evidence for more than a model?
Hi Tom,
Thanks for your suggestion. It’s nice to speak to you again. You have helped me a lot when I took the earlier Machine Learning course conducted via Matlab.