Bi-Level Optimization for Pseudo-Labeling Based Semi-Supervised Learning

Marzi Heidari, Yuhong Guo

Primary Area: unsupervised, self-supervised, semi-supervised, and supervised representation learning

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Keywords: Semi-Supervised Learning, Bi-level Optimization

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

Abstract:

Semi-supervised learning (SSL) is a fundamental task in machine learning, empowering models to extract valuable insights from datasets with limited labeled samples and a large amount of unlabeled data.
Although pseudo-labeling is a widely used approach for SSL that generates pseudo-labels for unlabeled data and leverages them as ground truth labels for training, traditional pseudo-labeling techniques often suffer from the problem of error accumulation, leading to a significant decrease in the quality of pseudo-labels and hence
the overall model performance.
In this paper, we propose a novel Bi-level Optimization method for Pseudo-label Learning (BOPL)
to boost semi-supervised training.
It treats pseudo-labels as latent variables, and optimizes the model parameters and pseudo-labels
jointly within a bi-level optimization framework.
By enabling direct optimization over the pseudo-labels towards maximizing the prediction model performance,
the method is expected to produce high-quality pseudo-labels that are much less susceptible to error accumulation.
To evaluate the effectiveness of the proposed approach,
we conduct extensive experiments on
multiple SSL benchmarks.
The experimental results show the proposed BOPL outperforms the state-of-the-art SSL techniques.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 4051