MBGDT:Robust Mini-Batch Gradient Descent

Hanming Wang, Haozheng Luo, Yue Wang

In high dimensions, most machine learning method perform fragile even there are a little outliers. To address this, we hope to introduce a new method with the base learner, such as…