Title: Conformal prediction of future insurance claims in the regression problem

Abstract: In the current insurance literature, prediction of insurance claims in the regression problem is often performed with a statistical model. This model-based approach may potentially suffer from several drawbacks: (i) model misspecification, (ii) selection effect, and (iii) lack of finite-sample validity. This article addresses these three issues simultaneously by employing conformal prediction-a general machine learning strategy for valid predictions. The proposed method is both model-free and tuning-parameter-free. It also guarantees finite-sample validity at a pre-assigned coverage probability level. Examples, based on both simulated and real data, are provided to demonstrate the excellent performance of the proposed method and its applications in insurance, especially regarding meeting the solvency capital requirement of European insurance regulation, Solvency II.

Title: Distributed and Online Methods in Quantile Regression

Abstract: Quantile regression offers a versatile framework for modeling heterogeneous effects, but modern big data present significant computational and methodological challenges. This talk will trace a line of research on distributed and online approaches for quantile regression and related problems. In this talk, I will discuss a sequence of developments in distributed and online approaches for quantile regression, drawing connections to both classical estimation strategies and more recent advances in causal inference. I will highlight key ideas, recent progress, and future directions, with an emphasis on algorithmic insights and their relevance for large-scale applications.

Dr. Adam Smercina, Space Telescope Science Institute

Title: A New Era of Galaxy Evolution using Resolved Stars

Abstract: The varied and dynamic evolutionary histories of galaxies give rise to their stunning diversity in the present-day universe. Inferring these histories requires accessing the information encoded in their longest-lived visible components: stars. We are in an exciting new frontier, with a fleet of current and upcoming observatories capable of accessing the resolved stellar populations within and around external galaxies. In this talk, I will first summarize my efforts to chart the merger histories of nearby galaxies by surveying the stars in their accreted halos, including the exciting potential of the upcoming Roman Space Telescope. I will then discuss my efforts to trace the evolution of these galaxies star formation and structure, particularly as a consequence of their merger histories, through high-resolution surveys of their main bodies. In particular, I will highlight several large programs with JWST, which has opened up an exciting new frontier for this science. Over the next decade, these efforts with JWST and Roman have the potential to transform our view of galaxy evolution. To close, I will discuss how this current pioneering work with JWST will pave way for the next paradigm shift in resolved star science: the Habitable Worlds Observatory.