Total: 1
We introduce statistical decision theory in stochastic thermodynamic systems. Statistical decision theory integrates several statistical tasks, such as estimation, hypothesis testing, and classification. In this study, we consider statistical decision theory in stochastic thermodynamics, which unifies sensing, classification, and uncertainty relations. In particular, we focus on the minimax risk, the smallest achievable expected loss under the worst-case parameter choice, even with an optimal estimator. We derive lower bounds on the minimax risk using Le Cam's two-point method. The derived relationships can be used across a broader range of thermodynamic scenarios. To illustrate these relations, we study classification and distribution estimation problems in a Markov process. We demonstrate that the minimax risk provides lower bounds given by the dynamical activity in these problems. Furthermore, by examining particular dynamics, we establish that the resulting bounds simplify to uncertainty relations involving dynamical activity or entropy production.