Summary of Ldacp: Long-delayed Ad Conversions Prediction Model For Bidding Strategy, by Peng Cui (1) et al.
LDACP: Long-Delayed Ad Conversions Prediction Model for Bidding Strategy
by Peng Cui, Yiming Yang, Fusheng Jin, Siyuan Tang, Yunli Wang, Fukang Yang, Yalong Jia, Qingpeng Cai, Fei Pan, Changcheng Li, Peng Jiang
First submitted to arxiv on: 25 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper proposes a Long-Delayed Ad Conversions Prediction model for bidding strategy (LDACP), which aims to predict the number of long-delayed conversions in online advertising. The current bidding system relies on real-time tracked conversions, but this can lead to conservative bidding strategies due to overestimation of the Cost Per Action (CPA). To address this challenge, the authors transform regression problems into bucket classification problems and propose two sub-modules: Bucket Classification Module with label Smoothing method (BCMS) and Value Regression Module with Proxy labels (VRMP). The BCMS converts one-hot hard labels into non-normalized soft labels to alleviate discontinuity issues, while VRMP uses prediction bias as proxy labels to predict tail data. A Mixture of Experts (MoE) structure integrates the predictions from both modules to obtain the final predicted ad conversion number. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper helps us better understand how online ads are priced and how we can make more accurate predictions about which ads will be successful. The authors created a new way to predict how many people will click on an ad after a while, rather than just looking at what’s happening right now. This is important because it means that the price of an ad can be adjusted in real-time to get the best results. |
Keywords
» Artificial intelligence » Classification » Mixture of experts » One hot » Regression
