arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2606.01596math.NAcs.LGcs.NA
收藏

通过二阶几何监督学习混沌动力学

Second-order consistency for learning chaotic dynamics via randomized Jacobian matching

Department of Computer Science and Software Engineering, Korea University(韩国大学计算机科学与软件工程系);Department of Aerospace Engineering and Engineering Mechanics, The Oden Institute for Computational Engineering and Sciences, The University of Texas at Austin(德克萨斯大学奥斯汀分校航空航天工程与工程力学系,奥登计算工程与科学研究所)

Shinhoo Kang, Hai V. Nguyen, Tan Bui-Thanh

AI总结:

提出模型约束随机雅可比匹配方法,以O(d^2)代价隐式施加二阶一致性,在混沌系统中恢复吸引子几何和不变统计量。

摘要:

从数据中学习混沌动力系统需要的不仅仅是短期预测精度:学习模型必须保持吸引子几何及其不变统计量。轨迹(零阶)和雅可比(一阶)匹配监督向量场的值和切结构,但两者都不约束场如何偏离其切平面。因此,模型可以在监督状态下匹配值和切线,但弯曲方式与真实情况不同,在保持局部精度的同时,向虚假吸引子漂移并扭曲长时间统计量。我们证明,强制二阶一致性可以减轻这些失败,但在高维中形成完整的Hessian矩阵是禁止的。我们提出模型约束随机雅可比匹配,该方法在随机扰动的输入处比较真实和学习的向量场的雅可比矩阵。泰勒展开表明,期望的随机雅可比损失分解为名义雅可比失配加上由噪声方差缩放的Hessian失配,从而以O(d^2)代价隐式施加二阶一致性,而无需形成O(d^3)的Hessian张量。仅使用雅可比评估,该方法可扩展到显式Hessian匹配无法实现的高维。数值实验证实二阶方法是稳健的。对于Lorenz~63,一阶方法在最小时间监督下产生灾难性的Lyapunov指数异常值,而二阶方法消除了这些异常值并恢复了正确的吸引子。对于耦合Lorenz~96,分布外强迫扫描区分了这些方法:所有方法在F=16之前一致,但超过F=18后,只有二阶方法保持了不变测度和Lyapunov谱。在两个系统上,随机雅可比匹配以低得多的成本实现了与显式Hessian匹配相当的性能。

英文摘要:

Short-horizon accuracy does not ensure that a learned chaotic system has correct long-time dynamics. Trajectory (zeroth-order) matching constrains vector-field values, and Jacobian (first-order) matching constrains local tangent dynamics, but neither determines how the Jacobian varies away from supervised states, so a model can be locally accurate while drifting toward spurious attractors and distorting long-time statistics. We show that second-order supervision mitigates these failures. Because forming full Hessian tensors is computationally prohibitive in high dimensions, we propose model-constrained randomized Jacobian matching, which compares the Jacobians of the true and learned vector fields at randomly perturbed inputs. A Taylor expansion shows that the expected randomized Jacobian loss decomposes into the Jacobian mismatch plus a Hessian mismatch scaled by the noise variance, implicitly enforcing second-order consistency at $O(d^2)$ memory cost without forming the $O(d^3)$ Hessian tensor. In Lorenz 63 with minimal temporal supervision, second-order supervision reduces invariant-measure error and Lyapunov-spectrum MSE, and recovers the constant Hessian norm of the true bilinear field. Across five training seeds, explicit Hessian matching produces catastrophic Lyapunov outliers for four of five seeds, whereas randomized Jacobian matching produces none among 5,000 on-attractor rollouts, and it attains the largest threshold in a directional capture scan. In coupled Lorenz 96, first-order methods enter spurious high-amplitude regimes as forcing increases, while second-order methods retain accurate marginals. Randomized Jacobian matching costs about the same as explicit Hessian matching on Lorenz 63 and 40% less on Lorenz 96, with no reference Hessian evaluations during training.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑