About this document

Sub-Sequential Physics-Informed Learning With State Space Model by woshishyhhb is a document available to read on EtoBox.

The document presents PINN-Mamba, a novel framework that utilizes State Space Models (SSM) to improve Physics-Informed Neural Networks (PINNs) for solving partial differential equations (PDEs). It addresses the issues of initial condition propagation and simplicity bias in traditional PINNs, achieving error reductions of up to 86.3% compared to existing architectures. The proposed method integrates sub-sequence modeling with SSM to enhance the accuracy and efficiency of numerical solutions to PDEs without r

Author

woshishyhhb

Language

EN