2D-OOB: Attributing Data Contribution Through Joint Valuation Framework
This addresses the need for fine-grained data valuation in machine learning, particularly for interpreting noisy data and security applications, though it is incremental as it builds on existing out-of-bag frameworks.
The paper tackles the problem of attributing data contribution at the cell level rather than per data point, proposing 2D-OOB to jointly value samples and cells, achieving state-of-the-art performance with exponential speed improvements in tasks like outlier detection and backdoor trigger localization.
Data valuation has emerged as a powerful framework for quantifying each datum's contribution to the training of a machine learning model. However, it is crucial to recognize that the quality of cells within a single data point can vary greatly in practice. For example, even in the case of an abnormal data point, not all cells are necessarily noisy. The single scalar score assigned by existing data valuation methods blurs the distinction between noisy and clean cells of a data point, making it challenging to interpret the data values. In this paper, we propose 2D-OOB, an out-of-bag estimation framework for jointly determining helpful (or detrimental) samples as well as the particular cells that drive them. Our comprehensive experiments demonstrate that 2D-OOB achieves state-of-the-art performance across multiple use cases while being exponentially faster. Specifically, 2D-OOB shows promising results in detecting and rectifying fine-grained outliers at the cell level, and localizing backdoor triggers in data poisoning attacks.