NTKCPL: Active Learning on Top of Self-Supervised Model by Estimating True Coverage
This work addresses the problem of high annotation costs in machine learning for researchers and practitioners by improving active learning with self-supervised models, though it is incremental as it builds on existing strategies.
The paper tackles the challenge of determining effective active learning strategies when combined with self-supervised models, where phase transitions occur earlier, by proposing NTKCPL, a method that estimates empirical risk using pseudo-labels and neural tangent kernel approximation, and it outperforms baselines on five datasets across a wider range of training budgets.
High annotation cost for training machine learning classifiers has driven extensive research in active learning and self-supervised learning. Recent research has shown that in the context of supervised learning different active learning strategies need to be applied at various stages of the training process to ensure improved performance over the random baseline. We refer to the point where the number of available annotations changes the suitable active learning strategy as the phase transition point. In this paper, we establish that when combining active learning with self-supervised models to achieve improved performance, the phase transition point occurs earlier. It becomes challenging to determine which strategy should be used for previously unseen datasets. We argue that existing active learning algorithms are heavily influenced by the phase transition because the empirical risk over the entire active learning pool estimated by these algorithms is inaccurate and influenced by the number of labeled samples. To address this issue, we propose a novel active learning strategy, neural tangent kernel clustering-pseudo-labels (NTKCPL). It estimates empirical risk based on pseudo-labels and the model prediction with NTK approximation. We analyze the factors affecting this approximation error and design a pseudo-label clustering generation method to reduce the approximation error. We validate our method on five datasets, empirically demonstrating that it outperforms the baseline methods in most cases and is valid over a wider range of training budgets.