Conformal Credal Self-Supervised Learning

Lienen, Julian ; Demir, Caglar und Hüllermeier, Eyke (September 2023): Conformal Credal Self-Supervised Learning. Twelfth Symposium on Conformal and Probabilistic Prediction with Applications (COPA 2023), Limassol, Cyprus, September 13-15, 2023. Papadopoulos, Harris; Nguyen, Khuong An; Boström, Henrik und Carlsson, Lars (eds.) : Vol. 204 PMLR. pp. 214-233

Abstract

In semi-supervised learning, the paradigm of self-training refers to the idea of learning from pseudo-labels suggested by the learner itself. Recently, corresponding methods have proven effective and achieve state-of-the-art performance, e.g., when applied to image classification problems. However, pseudo-labels typically stem from ad-hoc heuristics, relying on the quality of the predictions though without guaranteeing their validity. One such method, so-called credal self-supervised learning, maintains pseudo-supervision in the form of sets of (instead of single) probability distributions over labels, thereby allowing for a flexible yet uncertainty-aware labeling. Again, however, there is no justification beyond empirical effectiveness. To address this deficiency, we make use of conformal prediction, an approach that comes with guarantees on the validity of set-valued predictions. As a result, the construction of credal sets of labels is supported by a rigorous theoretical foundation, leading to better calibrated and less error-prone supervision for unlabeled data. Along with this, we present effective algorithms for learning from credal self-supervision. An empirical study demonstrates excellent calibration properties of the pseudo-supervision, as well as the competitiveness of our method on several image classification benchmark datasets.

• Item Type: Conference or Workshop Item (Paper)
• Faculties: Mathematics, Computer Science and Statistics > Computer Science > Artificial Intelligence and Machine Learning
• Subjects: 000 Computer science, information and general works > 000 Computer science, knowledge, and systems
• DFG: Gefördert durch die Deutsche Forschungsgemeinschaft (DFG) - 160364472