Dirichlet Active Learning
This work addresses active learning for scenarios with limited labeled data, particularly in graph-based domains, representing an incremental advancement with novel computational techniques.
The authors tackled the problem of active learning when labeled data is scarce by introducing Dirichlet Active Learning (DiAL), a Bayesian-inspired method that models feature-conditional class probabilities as a Dirichlet random field to improve calibration and enable classification and active learning, demonstrating competitiveness with state-of-the-art methods in low-label rate graph learning.
This work introduces Dirichlet Active Learning (DiAL), a Bayesian-inspired approach to the design of active learning algorithms. Our framework models feature-conditional class probabilities as a Dirichlet random field and lends observational strength between similar features in order to calibrate the random field. This random field can then be utilized in learning tasks: in particular, we can use current estimates of mean and variance to conduct classification and active learning in the context where labeled data is scarce. We demonstrate the applicability of this model to low-label rate graph learning by constructing ``propagation operators'' based upon the graph Laplacian, and offer computational studies demonstrating the method's competitiveness with the state of the art. Finally, we provide rigorous guarantees regarding the ability of this approach to ensure both exploration and exploitation, expressed respectively in terms of cluster exploration and increased attention to decision boundaries.