UAI 2023poster7 citations

Online estimation of similarity matrices with incomplete data

Fangchen Yu, Yicheng Zeng, Jianfeng Mao, Wenye Li

Abstract

The similarity matrix measures pairwise similarities between a set of data points and is an essential concept in data processing, routinely used in practical applications. Obtaining a similarity matrix is typically straightforward when data points are completely observed. However, incomplete observations can make it challenging to obtain a high-quality similarity matrix, which becomes even more complex in online data. To address this challenge, we propose matrix correction algorithms that leverage the positive semi-definiteness (PSD) of the similarity matrix to improve similarity estimation in both offline and online scenarios. Our approaches have a solid theoretical guarantee of performance and excellent potential for parallel execution on large-scale data. Empirical evaluations demonstrate their high effectiveness and efficiency with significantly improved results over classical imputation-based methods, benefiting downstream applications with superior performance. Our code is available at \url{https://github.com/CUHKSZ-Yu/OnMC}.

BibTeX
@InProceedings{pmlr-v216-yu23a,
  title = 	 {Online estimation of similarity matrices with incomplete data},
  author =       {Yu, Fangchen and Zeng, Yicheng and Mao, Jianfeng and Li, Wenye},
  booktitle = 	 {Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence},
  pages = 	 {2454--2464},
  year = 	 {2023},
  editor = 	 {Evans, Robin J. and Shpitser, Ilya},
  volume = 	 {216},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {31 Jul--04 Aug},
  publisher =    {PMLR},
  pdf = 	 {https://proceedings.mlr.press/v216/yu23a/yu23a.pdf},
  url = 	 {https://proceedings.mlr.press/v216/yu23a.html},
  abstract = 	 {The similarity matrix measures pairwise similarities between a set of data points and is an essential concept in data processing, routinely used in practical applications. Obtaining a similarity matrix is typically straightforward when data points are completely observed. However, incomplete observations can make it challenging to obtain a high-quality similarity matrix, which becomes even more complex in online data. To address this challenge, we propose matrix correction algorithms that leverage the positive semi-definiteness (PSD) of the similarity matrix to improve similarity estimation in both offline and online scenarios. Our approaches have a solid theoretical guarantee of performance and excellent potential for parallel execution on large-scale data. Empirical evaluations demonstrate their high effectiveness and efficiency with significantly improved results over classical imputation-based methods, benefiting downstream applications with superior performance. Our code is available at \url{https://github.com/CUHKSZ-Yu/OnMC}.}
}