Post Click

Post-click conversion rate (CVR) prediction aims to estimate the likelihood of a user completing a desired action (e.g., purchase) after clicking an advertisement or recommendation. Current research focuses on addressing challenges like data sparsity and selection bias using techniques such as multi-task learning models (e.g., Entire Space Multi-Task Models), self-supervised pre-training, and doubly robust learning frameworks to improve prediction accuracy and mitigate inherent biases. These advancements are crucial for optimizing online advertising, recommender systems, and other applications that rely on accurately predicting user behavior following initial engagement.