Replies: 1 comment
-
|
Hi @yspth23 , Thank you for contacting us! The decision to initially use time-invariant channel coefficients was primarily based on a trade-off between model complexity and stability. While time-varying coefficients can offer more flexibility, they significantly increase the number of parameters to be estimated. This can lead to issues with model identifiability and convergence, especially with limited data, as the model may struggle to learn meaningful signals from the noise. Our initial approach prioritized a simpler, more stable, and more widely applicable model. The Meridian team will consider incorporating time-varying coefficients as an optional feature as we work on future updates. You may track the changes to Meridian in the Change Log. Feel free to reach out if you have any questions or suggestions regarding Meridian. Thank you, Google Meridian Support Team |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
I understand that Meridian supports time varying intercept but the channel coefficients are assumed to be time invariant. Some other MMM frameworks/solutions do offer time varying channel parameters (and the business justification behind this generally looks plausible). So I assume there are some trade-offs or some reasons why Meridian authors decided against including them? I can't find info/discussion of this in the Meridian docs
Beta Was this translation helpful? Give feedback.
All reactions