Empirical Mode Modeling: A data-driven approach to recover and forecast nonlinear dynamics from noisy data
This incremental method addresses noise issues in state-space analysis for researchers studying complex systems like ecology or geophysics.
The paper tackles the problem of noise corrupting state-space representations in data-driven analysis of nonlinear systems, and finds that combining empirical mode decomposition with empirical dynamic modeling improves information content in noisy data, as shown in mathematical and geophysical applications.
Data-driven, model-free analytics are natural choices for discovery and forecasting of complex, nonlinear systems. Methods that operate in the system state-space require either an explicit multidimensional state-space, or, one approximated from available observations. Since observational data are frequently sampled with noise, it is possible that noise can corrupt the state-space representation degrading analytical performance. Here, we evaluate the synthesis of empirical mode decomposition with empirical dynamic modeling, which we term empirical mode modeling, to increase the information content of state-space representations in the presence of noise. Evaluation of a mathematical, and, an ecologically important geophysical application across three different state-space representations suggests that empirical mode modeling may be a useful technique for data-driven, model-free, state-space analysis in the presence of noise.