B1931
Title: Robust multi-task feature learning with adaptive Huber regressions
Authors: Xin Gao - York University (Canada) [presenting]
Abstract: When data from multiple tasks have outlier contamination, the performance of existing multi-task learning methods suffers efficiency loss. The robust multi-task featuring learning method is presented by combining the adaptive Huber regression tasks with mixed regularization. The robustification parameters can be chosen to adapt to the sample size, the model dimension, and the moments of the error distribution while striking a balance between unbiasedness and robustness. The method is shown to achieve estimation consistency and sign recovery consistency. In addition, the robust information criterion is proposed to conduct joint inference on related tasks, which can be used for consistent model selection. Simulation studies and real data analysis are provided to illustrate the performance of the proposed model.
