Principled model selection for stochastic dynamics

📅 2025-01-17
📈 Citations: 2
Influential: 0
📄 PDF
🤖 AI Summary
To address combinatorial explosion and overfitting arising from sparse library selection in stochastic differential equation (SDE) modeling, this paper proposes a parsimonious inference framework integrating likelihood-based statistics and extreme value theory (EVT). The method introduces EVT—novelly applied to SDE model selection—to identify extreme-event thresholds and perform statistical significance tests, thereby suppressing redundant parameters and enabling robust identification of minimal complete models. It unifies maximum likelihood estimation, sparse function library projection, and SDE discretization techniques. Experiments demonstrate that the framework maintains high accuracy under low sampling rates and strong measurement noise, significantly outperforming state-of-the-art methods in ecological network and reaction–diffusion system modeling. By bridging statistical rigor with physical interpretability, it establishes a new paradigm for interpretable modeling of complex stochastic dynamics.

Technology Category

Application Category

📝 Abstract
Complex dynamical systems, from macromolecules to ecosystems, are often modeled by stochastic differential equations. To learn such models from data, a common approach involves sparse selection among a large function library. However, we show that overfitting arises - not just from individual model complexity, but also from the combinatorial growth of possible models. To address this, we introduce Parsimonious Stochastic Inference (PASTIS), a principled method combining likelihood-estimation statistics with extreme value theory to suppress superfluous parameters. PASTIS outperforms existing methods and reliably identifies minimal models, even with low sampling rates or measurement error. It extends to stochastic partial differential equations, and applies to ecological networks and reaction-diffusion dynamics.
Problem

Research questions and friction points this paper is trying to address.

Addressing overfitting in stochastic differential equation model selection
Introducing a method to suppress superfluous parameters in dynamics models
Extending principled model selection to stochastic partial differential equations
Innovation

Methods, ideas, or system contributions that make the work stand out.

Parsimonious Stochastic Inference (PASTIS) method
Combines likelihood-estimation with extreme value theory
Suppresses superfluous parameters in model selection
🔎 Similar Papers
No similar papers found.
💼 Related Jobs
No related jobs found.
A
Andonis Gerardos
Aix Marseille Univ, CNRS, CINAM, Turing Center for Living Systems, Marseille, France
P
Pierre Ronceray
Aix Marseille Univ, CNRS, CINAM, Turing Center for Living Systems, Marseille, France