PulseAugur
LIVE 13:09:22
research · [2 sources] ·
0
research

ZNO introduces stable rational neural operators for discrete-time dynamics

Researchers have introduced the Z-Domain Neural Operator (ZNO), a novel causal neural operator designed for discrete-time dynamic systems. ZNO's architecture utilizes stable low-rank rational filters parameterized in the z-plane, addressing a gap in operator learning methods often focused on continuous-time problems. The model demonstrates particular effectiveness in system identification tasks involving stable rational systems with poles close to the unit circle and long memory dynamics. AI

Summary written by gemini-2.5-flash-lite from 2 sources. How we write summaries →

IMPACT Introduces a specialized neural operator for discrete-time system identification, potentially improving accuracy for specific dynamic systems.

RANK_REASON The cluster contains an arXiv preprint detailing a new model architecture for discrete-time dynamic systems.

Read on arXiv cs.LG →

COVERAGE [2]

  1. arXiv cs.LG TIER_1 · Xianli Zhu, Jia Yin ·

    ZNO: Stable Rational Neural Operators in the Z-Domain for Discrete-Time Dynamic

    arXiv:2605.02356v1 Announce Type: new Abstract: We introduce the Z-Domain Neural Operator (ZNO), a causal neural operator whose layers are stable low-rank multiple-input multiple-output (MIMO) rational filters parameterized directly in the $z$-plane. ZNO addresses a limitation of…

  2. arXiv cs.LG TIER_1 · Jia Yin ·

    ZNO: Stable Rational Neural Operators in the Z-Domain for Discrete-Time Dynamic

    We introduce the Z-Domain Neural Operator (ZNO), a causal neural operator whose layers are stable low-rank multiple-input multiple-output (MIMO) rational filters parameterized directly in the $z$-plane. ZNO addresses a limitation of existing operator learning methods, many of whi…