Data driven individualized decision making problems have received a lot of attentions in recent years. In particular, decision makers aim to determine the optimal Individualized Treatment Rule (ITR) so that the expected specified outcome averaging over heterogeneous patient-specific characteristics is maximized. Many existing methods deal with binary or a moderate number of treatment arms and may not take potential treatment effect structure into account. However, the effectiveness of these methods may deteriorate when the number of treatment arms becomes large. In this article, we propose GRoup Outcome Weighted Learning (GROWL) to estimate the latent structure in the treatment space and the optimal group-structured ITRs through a single optimization. In particular, for estimating group-structured ITRs, we utilize the Reinforced Angle based Multicategory Support Vector Machines (RAMSVM) to learn group-based decision rules under the weighted angle based multi-class classification framework. Fisher consistency, the excess risk bound, and the convergence rate of the value function are established to provide a theoretical guarantee for GROWL. Extensive empirical results in simulation studies and real data analysis demonstrate that GROWL enjoys better performance than several other existing methods.

Learning an optimal Individualized Treatment Rule (ITR) is a very important problem in precision medicine. This paper is concerned with the challenge when the number of treatment arms is large, and some groups of treatments in the large treatment space may work similarly for the patients. Motivated by the recent development of supervised clustering, we propose a novel adaptive fusion based method to cluster the treatments with similar treatment effects together and estimate the optimal ITR simultaneously through a single convex optimization. The problem is formulated as balancing \textit{loss}$+$\textit{penalty} terms with a tuning parameter, which allows the entire solution path of the treatment clustering process to be clearly visualized hierarchically. For computation, we propose an efficient algorithm based on accelerated proximal gradient and further conduct a novel group-lasso based algorithm for variable selection to boost the performance. Moreover, we demonstrate the theoretical guarantee of recovering the underlying true clustering structure of the treatments for our method. Finally, we demonstrate the superior performance of our method via both simulations and a real data application on cancer treatment, which may assist the decision making process for doctors.