Abstract: Despite the availability of large amounts of genomics data, medical treatment recommendations have yet to use them successfully. In this study, we consider the utility of high-dimensional genomic-clinical data and nonparametric methods for making cancer treatment recommendations. Our work builds on the framework of the individualized treatment rule (Qian and Murphy (2011)) but we aim to overcome their method's limitations, specifically when the method encounters a large number of covariates and when the model is misspecified. We tackle this problem using a dimension reduction method, namely Sliced Inverse Regression (SIR, Li (1991)), with a rich class of models for the treatment response. Notably, the SIR defines a feature space for high-dimensional data, offering an advantage similar to those found in the popular neural network models. With the features obtained from the SIR, we use a simple visualization to compare different treatment options and recommend a treatment. Additionally, we derive the consistency and the convergence rate of the proposed recommendation approach using a value function. Lastly, we demonstrate the effectiveness of the proposed approach using simulation studies and a real-data example of the treatment of multiple myeloma.

Key words and phrases: Dimension reduction, individualized treatment rules, sliced inverse regression, visualization.