CANCELED: Biostatistics Seminar Series: Min-ge Xie, PhD

Title: Individualized Fusion Learning (i-Fusion) for Individualized Inference   Abstract: Learnings from different data sources can often be fused together to yield more powerful findings than those from individual sources alone.  This talk presents a new fusion learning approach, named "i-Fusion", for drawing efficient individual inference by fusing leanings from relevant data sources. This i-Fusion is robust for handling heterogeneity arising from diverse sources in big data, and is ideally suited for goal-directed applications such as precision medicine. Specifically,  i-Fusion summarizes individual inference information in confidence distributions (CDs), then adaptively forms a clique of individuals that bear relevance to the target individual, and finally combines the CDs from those relevant individuals and draws inference for the target individual from the combined CD. In essence, i-Fusion "borrows strength" from relevant individuals to improve efficiency while retaining inference validity. Computationally, i-Fusion is parallel in nature and scales up well in comparison with existing competitors. Examples in simulations and real applications in financial forecasting are presented. (Joint work with Jieli Shen and Regina Liu)