Workshop SDS: Registration
View Submission - SDS2022
A0192
Title: (SAM)$^2$: A family of sequential sample averaging algorithms via majorization--minimization Authors:  Hien Nguyen - University of Queensland (Australia) [presenting]
Abstract: In practice, many data sets are retrieved sequentially or are too large to be stored in the memory of typical machines available to most analysts. We consider a sequential sampling approach for constructing optimization algorithms for many machine learning and statistical settings that can alleviate the difficulties caused by the aforementioned issues. The algorithms combine a sequential sampling scheme with a majorization--minimization approach to producing a globally convergent family of algorithms that has broad applicability.