Nonstationary Sparse Spectral Permanental Process
–Neural Information Processing Systems
Existing permanental processes often impose constraints on kernel types or sta-tionarity, limiting the model's expressiveness. To overcome these limitations, we propose a novel approach utilizing the sparse spectral representation of non-stationary kernels.
Neural Information Processing Systems
Feb-17-2026, 09:48:56 GMT
- Country:
- Asia > China
- Beijing > Beijing (0.04)
- Hubei Province (0.04)
- Asia > China
- Genre:
- Research Report > Experimental Study (0.93)
- Industry:
- Transportation (0.46)
- Technology: