Accelerating model convergence in resource-constrained environments is essential for fast and efficient neural network training. This work presents learn2mix, a new training strategy that adaptively adjusts class proportions within batches, focusing on classes with higher error rates. Unlike classical training methods that use static class proportions, learn2mix continually adapts class proportions during training, leading to faster convergence. Empirical evaluations on benchmark datasets show that neural networks trained with learn2mix converge faster than those trained with existing approaches, achieving improved results for classification, regression, and reconstruction tasks under limited training resources and with imbalanced classes. Our empirical findings are supported by theoretical analysis.

翻译：在资源受限环境下加速模型收敛对于实现快速高效的神经网络训练至关重要。本文提出Learn2Mix——一种通过自适应调整批次内类别比例、聚焦于高错误率类别的新型训练策略。与传统采用静态类别比例的训练方法不同，Learn2Mix在训练过程中持续调整类别比例，从而实现更快的收敛速度。在基准数据集上的实证评估表明，采用Learn2Mix训练的神经网络比现有方法收敛更快，在训练资源有限且类别不平衡的条件下，于分类、回归和重构任务中均取得了更优的结果。我们的实证发现得到了理论分析的支持。