Can I read Fast HOSVD for Volterra Kernels on EtoBox?

Fast HOSVD for Volterra Kernels by adam_annie is a document available to read on EtoBox.

What is Fast HOSVD for Volterra Kernels about?

This document proposes a fast orthogonal decomposition method for cubic Volterra kernels using oblique unfolding. Volterra models are used to represent nonlinear systems but have a large number of parameters. The multilinear SVD (HOSVD) can decompose Volterra kernels but has high computational cost. The proposed method reduces this cost by considering the symmetrized kernel and exploiting column redundancies in the unfolded matrix using an oblique unfolding, reducing the complexity from 12I^4 flops to 2I^4

Author

adam_annie

Language

EN