Learning Latent Memory States from Longitudinal Athlete Monitoring Data

Abstract

We propose a new unit of analysis for longitudinal data: the Latent Memory Table. The scientific contribution is not the encoder. It is that table, treated as a reusable statistical object. A memory operator maps each masked windowed history to a finite-dimensional state; collecting those states with uncertainty yields the Latent Memory Table. SoccerMon serves as an empirical case study.

Dae-Jin Lee

Assistant Professor in Statistical Sciences

P-splines, mixed models, computational statistics, flexible survival, and latent memory for athlete monitoring.