Probability Flow Approach to the Onsager--Machlup Functional for Jump-Diffusion Processes

Yuanfei Huang,Xiang Zhou,Jinqiao Duan

2024-09-02

Abstract:The Onsager--Machlup action functional is an important concept in statistical mechanics and thermodynamics to describe the probability of fluctuations in nonequilibrium systems. It provides a powerful tool for analyzing and predicting the behavior of complex stochastic systems. For diffusion process, the path integral method and the Girsanov transformation are two main approaches to construct the Onsager--Machlup functional. However, it is a long-standing challenge to apply these two methods to the jump-diffusion process, because the complexity of jump noise presents intrinsic technical barriers to derive the Onsager--Machlup functional. In this work, we propose a new strategy to solve this problem by utilizing the equivalent probabilistic flow between the pure diffusion process and the jump-diffusion process. For the first time, we rigorously establish the closed-form expression of the Onsager--Machlup functional for jump-diffusion processes with finite jump activity, which include an important term of the Lévy intensity at the origin. The same probability flow approach is further applied to the Lévy process with infinite jump activity, and yields a time-discrete version of the Onsager--Machlup functional.

Probability

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to derive the Onsager–Machlup functional of jump - diffusion processes. Specifically, the author proposes a new strategy. By using the probability - flow equivalence between pure - diffusion processes and jump - diffusion processes, the closed - form Onsager–Machlup functional of jump - diffusion processes with finite jump activities is strictly established for the first time. Moreover, for Lévy processes with infinite jump activities, this method also provides a time - discrete version of the Onsager–Machlup functional.

### Background and Challenges

1. **Limitations of Traditional Methods**:
- For pure - diffusion processes, the path - integral method and the Girsanov transformation are the two main methods for constructing the Onsager–Machlup functional.
- However, these methods encounter technical obstacles when applied to jump - diffusion processes because the complexity of jump noise makes it difficult to derive the Onsager–Machlup functional.

2. **Research Motivation**:
- The Onsager–Machlup functional is an important concept used in statistical mechanics and thermodynamics to describe the fluctuation probability in non - equilibrium systems.
- This functional provides a powerful tool for analyzing and predicting the behavior of complex stochastic systems.

### Main Contributions

1. **Probability - Flow Equivalence**:
- The author introduces the concept of probability - flow equivalence between pure - diffusion processes and jump - diffusion processes.
- Through this method, the closed - form Onsager–Machlup functional of jump - diffusion processes with finite jump activities is successfully derived.

2. **Specific Results**:
- For jump - diffusion processes with finite jump activities, the expression of the Onsager–Machlup functional is:
\[
S_{OM}^X(\psi, \dot{\psi})=\frac{1}{2} \int_{0}^{T}\left[\frac{1}{\sigma^{2}}\left|\dot{\psi}_{s}-b(\psi_{s})-\ell_{J}(\psi_{s})\right|^{2}+\nabla \cdot(b(\psi_{s})+\ell_{J}(\psi_{s}))+\tilde{\ell}_{J}(\psi_{s})\right] d s,
\]
where
\[
\ell_{J}(x):=\int_{\mathbb{R}^{d}} \int_{0}^{1} \lambda(x - \theta z)\frac{\nu_{J}(-\theta z)}{\nu_{J}(0)} d\theta \nu_{J}(d z),
\]
\[
\tilde{\ell}_{J}(x):=\int_{\mathbb{R}^{d}} \int_{0}^{1} \lambda(x - \theta z)\frac{\nabla \nu_{J}(-\theta z)\nu_{J}(0)-\nu_{J}(-\theta z)\nabla \nu_{J}(0)}{(\nu_{J}(0))^{2}} d\theta \nu_{J}(d z).
\]

3. **Extension to Infinite Jump Activities**:
- For jump - diffusion processes with infinite jump activities, the author provides a time - discrete version of the Onsager–Machlup functional.
- This discrete version can converge to the continuous - form Onsager–Machlup functional under specific conditions.

### Conclusion

This paper successfully solves the problem of deriving the Onsager–Machlup functional of jump - diffusion processes by introducing a new method of probability - flow equivalence, fills the theoretical gap, and provides a basis for further research in related fields.