Does the Definition of Difficulty Matter? Scoring Functions and their Role for Curriculum Learning

TL;DR

This study investigates how the definition of difficulty affects curriculum learning in deep models, analyzing scoring functions' robustness, their impact on training, and how different strategies can complement each other.

Contribution

It provides an extensive analysis of scoring functions for difficulty estimation, their robustness, and the effects of curriculum learning strategies across different tasks and settings.

Findings

Scoring functions are highly dependent on training randomness and can be stabilized with ensemble methods.

Curriculum learning's effectiveness varies; the data presentation order significantly influences performance.

Different curriculum strategies can complement each other, improving model accuracy through late fusion.

Abstract

Curriculum learning (CL) describes a machine learning training strategy in which samples are gradually introduced into the training process based on their difficulty. Despite a partially contradictory body of evidence in the literature, CL finds popularity in deep learning research due to its promise of leveraging human-inspired curricula to achieve higher model performance. Yet, the subjectivity and biases that follow any necessary definition of difficulty, especially for those found in orderings derived from models or training statistics, have rarely been investigated. To shed more light on the underlying unanswered questions, we conduct an extensive study on the robustness and similarity of the most common scoring functions for sample difficulty estimation, as well as their potential benefits in CL, using the popular benchmark dataset CIFAR-10 and the acoustic scene classification task from the DCASE2020 challenge as representatives of computer vision and computer audition, respectively. We report a strong dependence of scoring functions on the training setting, including randomness, which can partly be mitigated through ensemble scoring. While we do not find a general advantage of CL over uniform sampling, we observe that the ordering in which data is presented for CL-based training plays an important role in model performance. Furthermore, we find that the robustness of scoring functions across random seeds positively correlates with CL performance. Finally, we uncover that models trained with different CL strategies complement each other by boosting predictive power through late fusion, likely due to differences in the learnt concepts. Alongside our findings, we release the aucurriculum toolkit (https://github.com/autrainer/aucurriculum), implementing sample difficulty and CL-based training in a modular fashion.