Hello everyone,
I am looking to predict whether a user will buy a product on a website. The challenge is that users from over 87 different countries access the website, and I suspect that this heterogeneity may lead to poor predictions by the model. Should I use more than one model? Or should I use one model and tweak it in some way? Multilevel modeling? Dimensionality reduction?
Any advice would be greatly appreciated!
I am looking to predict whether a user will buy a product on a website. The challenge is that users from over 87 different countries access the website, and I suspect that this heterogeneity may lead to poor predictions by the model. Should I use more than one model? Or should I use one model and tweak it in some way? Multilevel modeling? Dimensionality reduction?
Any advice would be greatly appreciated!
