# MESA: Boost Ensemble Imbalanced Learning with MEta-SAmpler

arXiv: https://arxiv.org/abs/2010.08830 · HTML: https://mlsanity.com/paper/2010.08830/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Oct 2020.

## Built on by (most cited first, 25 of 25)

- [Data-centric Artificial Intelligence: A Survey](https://mlsanity.com/paper/2303.10158/) (2023)
- [Influence-Balanced Loss for Imbalanced Visual Classification](https://mlsanity.com/paper/2110.02444/) (2021)
- [ALIP: Adaptive Language-Image Pre-training with Synthetic Caption](https://mlsanity.com/paper/2308.08428/) (2023)
- [The Efficiency Spectrum of Large Language Models: An Algorithmic Survey](https://mlsanity.com/paper/2312.00678/) (2023)
- [Improved Multi-label Classification under Temporal Concept Drift: Rethinking Group-Robust Algorithms in a Label-Wise Setting](https://mlsanity.com/paper/2203.07856/) (2022)
- [Constrained Optimization to Train Neural Networks on Critical and Under-Represented Classes](https://mlsanity.com/paper/2102.12894/) (2021)
- [GraphSR: A Data Augmentation Algorithm for Imbalanced Node Classification](https://mlsanity.com/paper/2302.12814/) (2023)
- [SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidence](https://mlsanity.com/paper/2502.08767/) (2025)
- [Rethinking Semi-Supervised Imbalanced Node Classification from Bias-Variance Decomposition](https://mlsanity.com/paper/2310.18765/) (2023)
- [Fine-tune Language Models to Approximate Unbiased In-context Learning](https://mlsanity.com/paper/2310.03331/) (2023)
- [Breaking Silos: Adaptive Model Fusion Unlocks Better Time Series Forecasting](https://mlsanity.com/paper/2505.18442/) (2025)
- [A Comprehensive Survey on Imbalanced Data Learning](https://mlsanity.com/paper/2502.08960/) (2025)
- [Towards Automated Imbalanced Learning with Deep Hierarchical Reinforcement Learning](https://mlsanity.com/paper/2208.12433/) (2022)
- [AIM: Attributing, Interpreting, Mitigating Data Unfairness](https://mlsanity.com/paper/2406.08819/) (2024)
- [Data Optimization in Deep Learning: A Survey](https://mlsanity.com/paper/2310.16499/) (2023)
- [Training Class-Imbalanced Diffusion Model Via Overlap Optimization](https://mlsanity.com/paper/2402.10821/) (2024)
- [Handling Inter-class and Intra-class Imbalance in Class-imbalanced Learning](https://mlsanity.com/paper/2111.12791/) (2021)
- [Regularization Penalty Optimization for Addressing Data Quality Variance in OoD Algorithms](https://mlsanity.com/paper/2206.05749/) (2022)
- [Automated Imbalanced Learning](https://mlsanity.com/paper/2211.00376/) (2022)
- [BadSampler: Harnessing the Power of Catastrophic Forgetting to Poison Byzantine-robust Federated Learning](https://mlsanity.com/paper/2406.12222/) (2024)
- [Multi-granularity Relabeled Under-sampling Algorithm for Imbalanced Data](https://mlsanity.com/paper/2201.03957/) (2022)
- [Imbalanced Classification via Explicit Gradient Learning From Augmented Data](https://mlsanity.com/paper/2202.10550/) (2022)
- [ASE: Anomaly Scoring Based Ensemble Learning for Imbalanced Datasets](https://mlsanity.com/paper/2203.10769/) (2022)
- [Tackling Diverse Minorities in Imbalanced Classification](https://mlsanity.com/paper/2308.14838/) (2023)
- [Predicting the Reliability of an Image Classifier under Image Distortion](https://mlsanity.com/paper/2412.16881/) (2024)

## Abstract (arXiv)

Imbalanced learning (IL), i.e., learning unbiased models from class-imbalanced data, is a challenging problem. Typical IL methods including resampling and reweighting were designed based on some heuristic assumptions. They often suffer from unstable performance, poor applicability, and high computational cost in complex tasks where their assumptions do not hold. In this paper, we introduce a novel ensemble IL framework named MESA. It adaptively resamples the training set in iterations to get multiple classifiers and forms a cascade ensemble model. MESA directly learns the sampling strategy from data to optimize the final metric beyond following random heuristics. Moreover, unlike prevailing meta-learning-based IL solutions, we decouple the model-training and meta-training in MESA by independently train the meta-sampler over task-agnostic meta-data. This makes MESA generally applicable to most of the existing learning models and the meta-sampler can be efficiently applied to new tasks. Extensive experiments on both synthetic and real-world tasks demonstrate the effectiveness, robustness, and transferability of MESA. Our code is available at https://github.com/ZhiningLiu1998/mesa.