Machine learning constitutes model-building automation for data analysis and has achieved tremendous success in recent years. However, there is a lack of mathematical and statistical understanding of machine learning. Many machine learning methods, such as deep learning, are not as popular to statisticians as other methods. Hence, we aims to bring wider attention to developing machine learning and artificial intelligent based methods for statisticians and to bridge the gap between the theoretical understanding of using machine learning and statistical learning. Through a mix of applied and theoretical talks, five distinguished speakers from various domains including Industrial and Systems Engineering, Public Health, Business and Statistics, will showcase novel findings and cutting-edge achievements in statistical data analysis under the framework of machine learning. Four out of the five speakers are female statisticians with different career stages, including the Post-doctoral fellow, the associate professor and the full professors. The successful presentations of this invited session will significantly stimulate, encourage and strength the representation of women statisticians in the ISI and its Associations. In view of the variety of topics covered by the talks with emphasis on applications to fields including uncertainty quantification, neural networks, imaging analysis and topological and geometric statistics, the session will attract a wide audience from academia, industry and government.

In this session, five speakers will showcase novel findings and cutting-edge achievements in statistical data analysis under the framework of machine learning. Specifically, they will discuss the robust estimation in regression and classification methods for large dimensional data, optimal estimation under endogenous treatment assignment, robust sensible adversarial learning of deep neural networks for image classification. Last but not least, time zigzags at graph convolutional networks for time series forecasting and invertible neural networks for graph prediction will also be presented.

Organiser: Dr Guanqun Cao 

Chair: Dr Guanqun Cao 

Speaker: Dr Yao Xie 

Speaker: Yulia Gel 

Speaker: Jiwei Zhao  

Speaker: Dr Chunming Zhang  

Speaker:  Jungeun Kim