作者: V. Smelyanskiy , D. Luchinsky , D. Timuçin , A. Bandrivskyy
DOI: 10.1103/PHYSREVE.72.026202
关键词:
摘要: An algorithm is presented for reconstructing stochastic nonlinear dynamical models from noisy time-series data. The approach analytical; consequently, the resulting does not require an extensive global search model parameters, provides optimal compensation effects of noise, and robust a broad range models. strengths are illustrated by inferring parameters Lorenz system comparing results with those earlier research. efficiency accuracy further demonstrated five globally locally coupled oscillators.