Extended Data Fig. 1: The workflow of optimizing the parameters for the HMM.
From: Non-equilibrium entropy production and information dissipation in a non-Markovian quantum dot
Experimental data are split into train and test sets. Model parameters are optimized by minimizing the loss (top left) on the train data and validated by comparing losses on train and test sets. Bottom left: field-off parameters are optimized first and used for subsequent field-on optimization.
