Marketing Mix Optimization using Reinforcement Learning | Q-Learning & UCB | Python + FastAPI + n8n

Marketing Mix Optimization using Reinforcement Learning - A complete end-to-end system that learns optimal marketing budget allocation across channels using Q-Learning and UCB Bandit algorithms.

This project demonstrates:
Marketing Mix Model (MMM) with adstock and saturation effects
Q-Learning (Value-Based RL) for state-action optimization
UCB Contextual Bandit for exploration-exploitation balance
FastAPI REST endpoints for production deployment
n8n Agentic Workflow with natural language interface

Key Results:
MMM R² Score: 0.963
UCB outperforms Q-Learning by 6.5%
Both algorithms converge to ROI-Proportional allocation strategy

Tech Stack: Python, NumPy, Pandas, Scikit-learn, FastAPI, n8n, Google Gemini

Dataset: DT MART Market Mix Modeling (Kaggle)


GitHub Repository: [Add your link]

Timestamps:
0:00 Introduction
1:00 System Architecture
2:00 Training Demo
3:30 API Demo
4:30 n8n Agentic Demo
5:30 Results & Conclusion

#ReinforcementLearning #MachineLearning #MarketingAnalytics #Python #QLearning #UCB #FastAPI #n8n #AgenticAI #DataScience #MarketingMixModel #MMM #AI #DeepLearning #BudgetOptimization #MarTech #artificialintelligence

reinforcement learning, machine learning, marketing mix model, q-learning, ucb bandit, python project, fastapi, n8n workflow, agentic ai, marketing optimization, budget allocation, data science, artificial intelligence, mmm, marketing analytics, contextual bandit, exploration exploitation, martech, ai marketing, python tutorial