Behind the Scenes of an Ads Prediction System
- Track:
- PyData: Machine Learning, Stats
- Type:
- Talk
- Level:
- intermediate
- Room:
- Terrace 2B
- Start:
- 14:35 on 11 July 2024
- Duration:
- 30 minutes
Abstract
In this era of rapid technological advancement and AI, Ad prediction systems stand at the forefront of shaping online advertising, significantly impacting how content reaches its intended audience. In this session, I’ll introduce the Ads prediction system from a user and algorithm view. We’ll then walk through key concepts like targeting, bidding, ad ranking, click-through rate (CTR), and conversion rate. We’ll deeply dive into connecting the dots, designing an ad prediction system, some ethical considerations, models, offline and online metrics, scaling and deployment decisions that enable handling high volumes of data and requests efficiently, and some case studies. At the end of this session, attendees will comprehensively understand the end-to-end process of developing an ads prediction system.