Open kurvaraviteja355 opened 1 year ago
Yes, I'm referring to the same thing. For example, I have a dataset with various nations and multiple channels (some of which are hierarchical), and I need to determine the contribution of each channel and ROAS in each country. As an example, instead of one model for the country, develop a single hierarchical model. This also aids in learning from similar countries if any of the countries lacked sufficient data points.
@kurvaraviteja355 can you expand a bit on your request? We have an issue for adding hierarchical structure on advertisement channels https://github.com/pymc-labs/pymc-marketing/issues/88
Is that what you have in mind?