Collaborative Prediction: To Join or To Disjoin Datasets

Kyung Rok Kim, Yansong Wang, Xiaocheng Li, Guanting Chen
Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence, PMLR 286:2163-2201, 2025.

Abstract

With the recent rise of generative Artificial Intelligence (AI), the need of selecting high-quality dataset to improve machine learning models has garnered increasing attention. However, some part of this topic remains underexplored, even for simple prediction models. In this work, we study the problem of developing practical algorithms that select appropriate dataset to minimize population loss of our prediction model with high probability. Broadly speaking, we investigate when datasets from different sources can be effectively merged to enhance the predictive model’s performance, and propose a practical algorithm with theoretical guarantees. By leveraging an oracle inequality and data-driven estimators, the algorithm reduces population loss with high probability. Numerical experiments demonstrate its effectiveness in both standard linear regression and broader deep learning applications.

Cite this Paper

BibTeX
@InProceedings{pmlr-v286-kim25a, title = {Collaborative Prediction: To Join or To Disjoin Datasets}, author = {Kim, Kyung Rok and Wang, Yansong and Li, Xiaocheng and Chen, Guanting}, booktitle = {Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence}, pages = {2163--2201}, year = {2025}, editor = {Chiappa, Silvia and Magliacane, Sara}, volume = {286}, series = {Proceedings of Machine Learning Research}, month = {21--25 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v286/main/assets/kim25a/kim25a.pdf}, url = {https://proceedings.mlr.press/v286/kim25a.html}, abstract = {With the recent rise of generative Artificial Intelligence (AI), the need of selecting high-quality dataset to improve machine learning models has garnered increasing attention. However, some part of this topic remains underexplored, even for simple prediction models. In this work, we study the problem of developing practical algorithms that select appropriate dataset to minimize population loss of our prediction model with high probability. Broadly speaking, we investigate when datasets from different sources can be effectively merged to enhance the predictive model’s performance, and propose a practical algorithm with theoretical guarantees. By leveraging an oracle inequality and data-driven estimators, the algorithm reduces population loss with high probability. Numerical experiments demonstrate its effectiveness in both standard linear regression and broader deep learning applications.} }
Endnote
%0 Conference Paper %T Collaborative Prediction: To Join or To Disjoin Datasets %A Kyung Rok Kim %A Yansong Wang %A Xiaocheng Li %A Guanting Chen %B Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2025 %E Silvia Chiappa %E Sara Magliacane %F pmlr-v286-kim25a %I PMLR %P 2163--2201 %U https://proceedings.mlr.press/v286/kim25a.html %V 286 %X With the recent rise of generative Artificial Intelligence (AI), the need of selecting high-quality dataset to improve machine learning models has garnered increasing attention. However, some part of this topic remains underexplored, even for simple prediction models. In this work, we study the problem of developing practical algorithms that select appropriate dataset to minimize population loss of our prediction model with high probability. Broadly speaking, we investigate when datasets from different sources can be effectively merged to enhance the predictive model’s performance, and propose a practical algorithm with theoretical guarantees. By leveraging an oracle inequality and data-driven estimators, the algorithm reduces population loss with high probability. Numerical experiments demonstrate its effectiveness in both standard linear regression and broader deep learning applications.
APA
Kim, K.R., Wang, Y., Li, X. & Chen, G.. (2025). Collaborative Prediction: To Join or To Disjoin Datasets. Proceedings of the Forty-first Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 286:2163-2201 Available from https://proceedings.mlr.press/v286/kim25a.html.

Related Material