How Sensitive are Meta-Learners to Dataset Imbalance?Download PDF

13 Mar 2021 (modified: 22 Oct 2023)Learning to Learn 2021Readers: Everyone
Keywords: meta-learning, few-shot-learning, dataset imbalance
TL;DR: We show that meta-learners are quite robust to meta-dataset imbalance compared to task-level imbalance.
Abstract: Meta-Learning (ML) has proven to be a useful tool for training Few-Shot Learning (FSL) algorithms by exposure to batches of tasks sampled from a meta-dataset. However, the standard training procedure overlooks the dynamic nature of the real-world where object classes are likely to occur at different frequencies. While it is generally understood that imbalanced tasks harm the performance of supervised methods, there is no significant research examining the impact of imbalanced meta-datasets on the FSL evaluation task. This study exposes the magnitude and extent of this problem. Our results show that ML methods are more robust against meta-dataset imbalance than imbalance at the task-level with a similar imbalance ratio ($\rho<20$), with the effect holding even in long-tail datasets under a larger imbalance ($\rho=65$). Overall, these results highlight an implicit strength of ML algorithms, capable of learning generalizable features under dataset imbalance and domain-shift. The code to reproduce the experiments is released under an open-source license.
Proposed Reviewers: Mateusz Ochal, m.ochal@hw.ac.uk Massimiliano Patacchiola, mpatacch@ed.ac.uk
Community Implementations: [![CatalyzeX](/images/catalyzex_icon.svg) 1 code implementation](https://www.catalyzex.com/paper/arxiv:2104.05344/code)
0 Replies

Loading