Replication package for "A Kalman–Bucy–Koopman Filter for Joint State and Deterioration-Rate Estimation in Partially Observed Production–Inventory Systems"

Description

Replication package for the paper “A Kalman–Bucy–Koopman Filter for Joint State and Deterioration-Rate Estimation in Partially Observed Production–Inventory Systems” by Lotfi Tadj and Boumediene Hamzi. This archive contains the Python testbed and derived data needed to reproduce every figure, table, and theoretical verification in the paper. The code implements a deteriorating-inventory model with stock-dependent demand, a kernel/RKHS construction of the principal Koopman eigenfunctions that remains valid at the parameter-augmentation center eigenvalue, and the resulting joint Kalman–Bucy–Koopman filter together with EKF/UKF baselines. Contents code/ — model definition, kernel-eigenfunction construction, the joint KBK/EKF/UKF filters, and the scripts for each experiment (Monte-Carlo joint estimation, multi-store sweep, time-varying deterioration drift, closed-loop MPC, decision-relevance, and the Theorem-2/F1 curvature verification). The README maps each figure and table to its script. data/ — the demand series that drives all experiments: daily unit sales for a single store and perishable family (store 44, PRODUCE), cleaned to…

Description from Zenodo, shortened.

Files (1)

• ZIPKBK_replication_zenodo.zip91 kB

DataHunter does not store or host these files: every link leads to the original source. Check the license there before using the data.