[CS] TODAY: Zixin Ding MS PresentationMay 27, 2025

[via cs]

This is an announcement of Zixin Ding's MS Presentation
===============================================
Candidate: Zixin Ding

Date: Tuesday, May 27, 2025

Time:  1 pm CST

Remote Location: https://uchicago.zoom.us/j/3185473352?pwd=THgzTDVYKzhDcUZQU3AzS1pkMjI4UT09

Location: JCL 354

Title: Learning to rank for active learning via multi-task bilevel optimization

Abstract: Active learning is a promising paradigm to reduce the labeling cost by strategically requesting labels to improve model performance. However, existing active learning methods often rely on expensive acquisition function to compute, extensive model- ing retraining and multiple rounds of interaction with annotators. To address these limitations, we propose a novel approach for active learning, which aims to select batches of unlabeled instances through a learned surrogate model for data acquisi- tion. A key challenge in this approach is developing an acquisition function that generalizes well, as the history of data, which forms part of the utility function’s input, grows over time. Our novel algorithmic contribution is a bilevel multi-task bilevel optimization framework that predicts the relative utility—measured by the validation accuracy—of different training sets, and ensures the learned acquisition function generalizes effectively. For cases where validation accuracy is expensive to evaluate, we introduce efficient interpolation-based surrogate models to estimate the utility function, reducing the evaluation cost. We demonstrate the performance of our approach through extensive experiments on standard active classification benchmarks. By employing our learned utility function, we show significant im- provements over traditional techniques, paving the way for more efficient and effective utility maximization in active learning applications.


Advisors: Yuxin Chen

Committee Members:  Yuxin Chen, Haifeng Xu, Ruoxi Jia