lightweight_mmm
lightweight_mmm copied to clipboard
google
LightweightMMM 🦇 is a lightweight Bayesian Marketing Mix Modeling (MMM) library that allows users to easily train MMMs and obtain channel attribution information.
Can someone please explain what is a good number for these metrics (if anything called be "good")? Also, how would I explain it to someone from the business side who...
Hi When I run budget allocator, this returns an incorrect allocation; the budget for the two channels I used is around 50%; 50%, but somehow lightweight only gives 2% allocation...
Using nested models is often prescribed for mitigating search spend being endogenous variable. I read this [Google paper](https://arxiv.org/abs/1807.03292)(Bias Correction For Paid Search), It is mentioned we can use google query...
When trying to do this: media_scaler = preprocessing.CustomScaler(divide_operation=jnp.mean) target_scaler = preprocessing.CustomScaler(divide_operation=jnp.mean) cost_scaler = preprocessing.CustomScaler(divide_operation=jnp.mean) media_data_train = media_scaler.fit_transform(media_data_train) target_train = target_scaler.fit_transform(target_train) costs2 = cost_scaler.fit_transform(costs) I got the error. up until then...
I understand that model type - carryover only takes into the consideration the delayed lagged effect (peaks observed can have delay in it) , however it does not take into...
The pull request implements a new variant of the model("hill_carryover") which applies saturation on top of carryover formulation of model. This implementation aligns with the paper [([Jin, Y., Wang, Y.,...
Hi Team, just looking at the repo, the only reason i can see that we are including tensorflow as a requirement is to use the Gfile function in the utils.py...
Hello We are running lightweightMMM in a Databricks environment using a GPU runtime and we’ve noticed that GPU performance is much slower (10x) than running using CPUs. We are trying...
Hi all, I am trying to build LMMM using US states as geos. Currently, model provides functions for channel attribution and channel response curves at the overall level. Is it...
I am using hill_adstock LMMM with weekly spend. I have some questions about it. Can someone knowledgeable please answer these? 1) The response curves that are returned, which half-saturation and...