A0466
Title: Sparse partial generalized tensor regression
Authors: Dayu Sun - Indiana University (United States) [presenting]
Abstract: Tensor data, often characterized as multi-dimensional arrays, have become increasingly prevalent in biomedical studies, particularly in neuroimaging applications. Analyzing these complex datasets can be challenging due to the high dimensionality and inherent structures within tensors. The sparse partial generalized tensor regression (SPGTR) method is proposed for modeling general types of outcomes involving both tensor and vector/scalar predictors. The novel mode-wise penalized manifold optimization techniques enable the achievement of dimension reduction and sparsity in tensor coefficient estimation, improving the overall prediction performance. The asymptotic behavior of the proposed estimation is established. It demonstrates the effectiveness of the SPGTR through extensive simulation studies, and its application is showcased in investigating the association between posttraumatic stress disorder (PTSD) and brain connectivity matrices derived from functional magnetic resonance imaging (fMRI) data.
