Hardik Singh

Text Enrichment Pipeline

Supply Chain Operations Analytics

Contextual Bandit Reinforcement Learning for Ad Click Optimization

Developed a Contextual Bandit model using the Avazu Click-Through Rate dataset to optimize advertisement selection based on user context. Evaluated policy performance using Inverse Propensity Scoring (IPS) and Weighted Importance Sampling (WIS), demonstrating practical reinforcement learning techniques for sequential decision-making.