Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
DY: Fachverband Dynamik und Statistische Physik
DY 43: Stochastic thermodynamics and information processing
DY 43.10: Vortrag
Mittwoch, 14. März 2018, 12:30–12:45, BH-N 243
Testing optimality of sequential decision-making — •Izaak Neri1,2,5, Meik Dörpinghaus3,5, Édgar Roldán1,5, Heinrich Meyr4,5, and Frank Jülicher1,5 — 1Max Planck Institute for the Physics of Complex Systems, Nöthnitzer Straße 38, 01187 Dresden, Germany — 2Max Planck Institute of Molecular Cell Biology and Genetics, Pfotenhauerstraße 108, 01307 Dresden, Germany — 3Vodafone Chair Mobile Communications Systems, Technische Universität Dresden, 01062 Dresden, Germany — 4Institute for Integrated Signal Processing Systems, RWTH Aachen University, 52056 Aachen, Germany — 5Center for Advancing Electronics Dresden, 01062 cfaed, Germany
Sequential decision making studies how a device makes as fast as possible decisions based on the sequential observation of a stochastic process. Here, a statistical test for the optimality of black box decision devices and a related measure for the divergence to optimality of black box decision devices are presented; a device is optimal if the average decision time -- over different realisations of the observed process -- is minimal given a certain maximal allowed error probability. A relation between the presented statistical test and the recently derived fluctuation theorems for first-passage times of entropy production is discussed [1]. Numerical experiments illustrate the use of the test for practical purposes.
[1]I Neri, É Roldán, F Jülicher, Statistics of Infima and Stopping Times of Entropy Production and Applications to Active Molecular Processes, Physical Review X 7, 011019 2017