A Unifying Probabilistic Framework for Partially Labeled Data Learning

Partially labeled data learning (PLDL), including partial label learning (PLL) and partial multi-label learning (PML), has been widely used in nowadays data science. Researchers attempt to construct different specific models to deal with the different classification tasks for PLL and PML scenarios respectively. The main challenge in training classifiers for PLL and PML is how to deal with ambiguities caused by the noisy false-positive labels in the candidate label set. The state-of-the-art strategy for both scenarios is to perform disambiguation by identifying the ground-truth label(s) directly from the candidate label set, which can be summarized into two categories: 'the identifying method' and 'the embedding method'. However, both kinds of methods are constructed by hand-designed heuristic modeling under considerations like feature/label correlations with no theoretical interpretation. Instead of adopting heuristic or specific modeling, we propose a novel unifying framework called A Unifying Probabilistic Framework for Partially Labeled Data Learning (UPF-PLDL), which is derived from a clear probabilistic formulation, and brings existing research on PLL and PML under one theoretical interpretation with respect to information theory. Furthermore, the proposed UPF-PLDL also unifies 'the identifying method' and 'the embedding method' into one integrated framework, which naturally incorporates the feature and label correlation considerations. Comprehensive experiments on synthetic and real-world datasets for both PLL and PML scenarios clearly demonstrate the superiorities of the derived framework.

Automated summary

Partially labeled data learning is widely used in data science, but the challenge lies in handling ambiguities caused by false-positive labels. The current strategy is to identify the ground-truth labels from the candidate set, but it lacks theoretical interpretation. Instead, we propose a novel unifying probabilistic framework that provides a clear formulation and theoretical interpretation for PLL and PML. Our framework also integrates the identifying and embedding methods, considering feature and label correlations. Experimental results show the superiority of our derived framework in both PLL and PML scenarios.