Researchers have developed fTNN, a deterministic tensor neural network designed to solve fractional partial differential equations (PDEs). This method employs a geometry-adapted integration split and specialized quadrature techniques to handle the fractional Laplacian operator. The framework is particularly effective for problems with strong boundary singularities and long-time simulations, showing improved accuracy over existing fPINN and Monte Carlo baselines. AI
IMPACT Introduces a novel neural network architecture for solving complex mathematical problems, potentially advancing scientific computing.
RANK_REASON The cluster describes a new research paper detailing a novel method for solving fractional PDEs using a tensor neural network.
- alphaXiv
- CatalyzeX Code Finder for Papers
- CORE Recommender
- DagsHub
- FPInnovations
- Fractional Laplacian
- fractional Poisson equation
- fTNN
- Gauss-Jacobi quadrature
- Gotit.pub
- Hugging Face
- Influence Flower
- Monte Carlo
- ScienceCast
- time-dependent fractional advection-diffusion equation
AI-generated summary · Google Gemini · from 3 sources. How we write summaries →
