Confidence-Aware Multi-Field Model Calibration
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866929350304071680 |
|---|---|
| author | Zhao, Yuang Wu, Chuhan Jia, Qinglin Zhu, Hong Yan, Jia Zong, Libin Zhang, Linxuan Dong, Zhenhua Zhang, Muyu |
| author_facet | Zhao, Yuang Wu, Chuhan Jia, Qinglin Zhu, Hong Yan, Jia Zong, Libin Zhang, Linxuan Dong, Zhenhua Zhang, Muyu |
| contents | Accurately predicting the probabilities of user feedback, such as clicks and conversions, is critical for advertisement ranking and bidding. However, there often exist unwanted mismatches between predicted probabilities and true likelihoods due to the rapid shift of data distributions and intrinsic model biases. Calibration aims to address this issue by post-processing model predictions, and field-aware calibration can adjust model output on different feature field values to satisfy fine-grained advertising demands. Unfortunately, the observed samples corresponding to certain field values can be seriously limited to make confident calibrations, which may yield bias amplification and online disturbance. In this paper, we propose a confidence-aware multi-field calibration method, which adaptively adjusts the calibration intensity based on confidence levels derived from sample statistics. It also utilizes multiple fields for joint model calibration according to their importance to mitigate the impact of data sparsity on a single field. Extensive offline and online experiments show the superiority of our method in boosting advertising performance and reducing prediction deviations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_17655 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Confidence-Aware Multi-Field Model Calibration Zhao, Yuang Wu, Chuhan Jia, Qinglin Zhu, Hong Yan, Jia Zong, Libin Zhang, Linxuan Dong, Zhenhua Zhang, Muyu Machine Learning Accurately predicting the probabilities of user feedback, such as clicks and conversions, is critical for advertisement ranking and bidding. However, there often exist unwanted mismatches between predicted probabilities and true likelihoods due to the rapid shift of data distributions and intrinsic model biases. Calibration aims to address this issue by post-processing model predictions, and field-aware calibration can adjust model output on different feature field values to satisfy fine-grained advertising demands. Unfortunately, the observed samples corresponding to certain field values can be seriously limited to make confident calibrations, which may yield bias amplification and online disturbance. In this paper, we propose a confidence-aware multi-field calibration method, which adaptively adjusts the calibration intensity based on confidence levels derived from sample statistics. It also utilizes multiple fields for joint model calibration according to their importance to mitigate the impact of data sparsity on a single field. Extensive offline and online experiments show the superiority of our method in boosting advertising performance and reducing prediction deviations. |
| title | Confidence-Aware Multi-Field Model Calibration |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2402.17655 |