Exploring the Impact of Dataset Bias on Dataset Distillation
This work addresses the problem of dataset bias in dataset distillation for machine learning practitioners, but it is incremental as it explores an overlooked aspect without introducing a new method.
The paper investigates how dataset bias affects dataset distillation, finding that biases in the original dataset significantly degrade the performance of synthetic datasets in most cases, highlighting the need to address bias in this process.
Dataset Distillation (DD) is a promising technique to synthesize a smaller dataset that preserves essential information from the original dataset. This synthetic dataset can serve as a substitute for the original large-scale one, and help alleviate the training workload. However, current DD methods typically operate under the assumption that the dataset is unbiased, overlooking potential bias issues within the dataset itself. To fill in this blank, we systematically investigate the influence of dataset bias on DD. To the best of our knowledge, this is the first exploration in the DD domain. Given that there are no suitable biased datasets for DD, we first construct two biased datasets, CMNIST-DD and CCIFAR10-DD, to establish a foundation for subsequent analysis. Then we utilize existing DD methods to generate synthetic datasets on CMNIST-DD and CCIFAR10-DD, and evaluate their performance following the standard process. Experiments demonstrate that biases present in the original dataset significantly impact the performance of the synthetic dataset in most cases, which highlights the necessity of identifying and mitigating biases in the original datasets during DD. Finally, we reformulate DD within the context of a biased dataset. Our code along with biased datasets are available at https://github.com/yaolu-zjut/Biased-DD.