DeepCausalMMM

Overview

A Deep Learning framework for Marketing Mix Modeling with causal structure learning. Standard MMM approaches rely on linear regression or Bayesian hierarchical models that assume channel independence and struggle to capture temporal dynamics and non-linear saturation. DeepCausalMMM combines GRU-based temporal modeling with DAG-based causal discovery to recover channel relationships and time-varying effects that those approaches miss.

Methodology

Paper, code, and docs

How to cite

@article{PuttaparthiTirumala2026,
  author  = {Puttaparthi Tirumala, Aditya},
  title   = {DeepCausalMMM: A Deep Learning Framework for Marketing Mix Modeling with Causal Structure Learning},
  journal = {Journal of Open Source Software},
  year    = {2026},
  volume  = {11},
  number  = {120},
  pages   = {9914},
  doi     = {10.21105/joss.09914},
  url     = {https://doi.org/10.21105/joss.09914}
}

Independent discussions

Technical & practitioner coverage

Research & technical discovery