Changjian Shui

TL;DR: A unified and principled method for both the querying and training processes in deep batch active learning is proposed, providing theoretical insights from the intuition of modeling the interactive procedure in active learning as distribution matching by adopting the Wasserstein distance.

TL;DR: An upper bound on the generalization error of multitask learning is provided, showing the benefit of explicit and implicit task similarity knowledge, and a new training algorithm is proposed to learn the task relation coefficients and neural network parameters iteratively.

TL;DR: This work tackles the domain generalization problem to learn from multiple source domains and generalize to a target domain with unknown statistics and adopts optimal transport with Wasserstein distance, which could constrain the class label similarity, for adversarial training and also further deploy a metric learning objective to leverage the label information for achieving distinguishable classification boundary.

TL;DR: The CSP is reformulates as a constrained minimization problem and the equivalence of the reformulated and the original CSPs is established, which validate the efficiency and effectiveness of the proposed CSP formulation in different learning contexts.

TL;DR: This work tackles the domain generalization problem to learn from multiple source domains and generalize to a target domain with unknown statistics and adopts optimal transport with Wasserstein distance, which could constrain the class label similarity, for adversarial training and also further deploy a metric learning objective to leverage the label information for achieving distinguishable classification boundary.