Isotonic subgroup selection and its role in clinical trials

Richard Samworth

Cambridge

Friday 31st May 2024

14:30:00

B.51

Given a sample of covariate-response pairs, we consider the subgroup selection problem of identifying a subset of the covariate domain where the regression function exceeds a pre-determined threshold. This is particularly relevant in clinical trials and other applications in the life sciences, for instance where we might want to determine a subgroup that would benefit from a vaccine or treatment. We introduce a computationally-feasible approach for subgroup selection in the context of multivariate isotonic regression based on martingale tests and multiple testing procedures for logically-structured hypotheses. Our proposed procedure satisfies a non-asymptotic, uniform Type I error rate guarantee with power that attains the minimax optimal rate up to poly-logarithmic factors. Extensions cover classification, isotonic quantile regression and heterogeneous treatment effect settings. Questa iniziativa è parte delle Ph.D. Lectures finanziate dal Dipartimento di Eccellenza, 2023-2027, Dipartimento di Milano, Politecnico di Milano. La Lezione è corredata da incontri con gli studenti di dottorato, per approfondimenti legati al tema di ricerca presentato nella lezione, ed alle ricerche specifiche dei singoli studenti. Contatto: laura.sangalli@polimi.it.