通过稀疏变分高斯过程 Kolmogorov-Arnold 网络实现可扩展且可解释的科学发现(SVGP KAN)
Scalable and Interpretable Scientific Discovery via Sparse Variational Gaussian Process Kolmogorov-Arnold Networks (SVGP KAN)
AI总结:
SVGP-KAN 通过稀疏变分推断和 KAN 结构,实现可扩展且可解释的概率 KAN,适用于大规模科学数据集,并能识别相关输入和功能关系。
摘要:
Kolmogorov-Arnold 网络(KAN)通过在网络边放置可学习的单变量函数,为多层感知机(MLP)提供了一种有前景的替代方案,从而增强了可解释性。然而,标准 KAN 缺乏概率输出,限制了其在需要不确定性量化应用中的实用性。尽管最近的高斯过程(GP)扩展解决了这一问题,但它们使用精确推断方法,计算复杂度与数据大小 N 成立方关系,限制了其在小数据集上的应用。我们引入了稀疏变分 GP-KAN(SVGP-KAN),该架构将稀疏变分推断与 KAN 结构相结合。通过使用 M 个诱导点和解析矩匹配,我们的方法将计算复杂度从 O(N³) 降低到 O(NM²) 或与样本大小线性相关,使概率 KAN 能够应用于更大的科学数据集。此外,我们证明了通过整合基于排列的重要性分析,网络可以作为结构识别框架,识别相关输入并分类函数关系。
英文摘要:
Kolmogorov-Arnold Networks (KANs) offer a promising alternative to Multi-Layer Perceptron (MLP) by placing learnable univariate functions on network edges, enhancing interpretability. However, standard KANs lack probabilistic outputs, limiting their utility in applications requiring uncertainty quantification. While recent Gaussian Process (GP) extensions to KANs address this, they utilize exact inference methods that scale cubically with data size N, restricting their application to smaller datasets. We introduce the Sparse Variational GP-KAN (SVGP-KAN), an architecture that integrates sparse variational inference with the KAN topology. By employing $M$ inducing points and analytic moment matching, our method reduces computational complexity from $O(N^3)$ to $O(NM^2)$ or linear in sample size, enabling the application of probabilistic KANs to larger scientific datasets. Furthermore, we demonstrate that integrating a permutation-based importance analysis enables the network to function as a framework for structural identification, identifying relevant inputs and classifying functional relationships.
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。