visaOTR

Valid Improved Sparsity A-Learning (VISA) provides a new method for selecting important variables involved in optimal treatment regime from a multiply robust perspective. The VISA estimator achieves its success by borrowing the strengths of both model averaging (ARM, Yuhong Yang, 2001) doi:10.1198/016214501753168262 and variable selection (PAL, Chengchun Shi, Ailin Fan, Rui Song and Wenbin Lu, 2018) doi:10.1214/17-AOS1570. The package is an implementation of Zishu Zhan and Jingxiao Zhang. (2022+).