Title: Multi-category individualized treatment regime
Authors: Jin Xu - East China Normal University (China) [presenting]
Xinyang Huang - East China Normal University (China)
Abstract: Individualized treatment regimes (ITRs) aim to recommend treatments based on patient-specific characteristics in order to maximize the expected clinical outcome. Outcome weighted learning approaches have been proposed for this optimization problem with a primary focus on the binary treatment case. We propose a general framework for multi-category ITRs using generic surrogate risk. The proposed method accommodates the situations when the outcome takes a negative value and/or when the propensity score is unknown. At the same time, risks caused by different adverse events cannot be ignored. We thus propose another method to estimate an optimal individualized treatment rule that maximizes clinical benefit outcome while having the risk-controlled at the desired level. The proposed procedure employs a novel surrogate loss and a relaxed difference of convex functions algorithm to solve the nonconvex constrained optimization problem. Simulations and real data examples are used to demonstrate the finite sample performance.