Saved in:
Bibliographic Details
Main Authors: Yu, Annan, Mahoney, Michael W., Erichson, N. Benjamin
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.13975
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910627887316992
author Yu, Annan
Mahoney, Michael W.
Erichson, N. Benjamin
author_facet Yu, Annan
Mahoney, Michael W.
Erichson, N. Benjamin
contents State-space models (SSMs) that utilize linear, time-invariant (LTI) systems are known for their effectiveness in learning long sequences. To achieve state-of-the-art performance, an SSM often needs a specifically designed initialization, and the training of state matrices is on a logarithmic scale with a very small learning rate. To understand these choices from a unified perspective, we view SSMs through the lens of Hankel operator theory. Building upon it, we develop a new parameterization scheme, called HOPE, for LTI systems that utilizes Markov parameters within Hankel operators. Our approach helps improve the initialization and training stability, leading to a more robust parameterization. We efficiently implement these innovations by nonuniformly sampling the transfer functions of LTI systems, and they require fewer parameters compared to canonical SSMs. When benchmarked against HiPPO-initialized models such as S4 and S4D, an SSM parameterized by Hankel operators demonstrates improved performance on Long-Range Arena (LRA) tasks. Moreover, our new parameterization endows the SSM with non-decaying memory within a fixed time window, which is empirically corroborated by a sequential CIFAR-10 task with padded noise.
format Preprint
id arxiv_https___arxiv_org_abs_2405_13975
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle HOPE for a Robust Parameterization of Long-memory State Space Models
Yu, Annan
Mahoney, Michael W.
Erichson, N. Benjamin
Machine Learning
State-space models (SSMs) that utilize linear, time-invariant (LTI) systems are known for their effectiveness in learning long sequences. To achieve state-of-the-art performance, an SSM often needs a specifically designed initialization, and the training of state matrices is on a logarithmic scale with a very small learning rate. To understand these choices from a unified perspective, we view SSMs through the lens of Hankel operator theory. Building upon it, we develop a new parameterization scheme, called HOPE, for LTI systems that utilizes Markov parameters within Hankel operators. Our approach helps improve the initialization and training stability, leading to a more robust parameterization. We efficiently implement these innovations by nonuniformly sampling the transfer functions of LTI systems, and they require fewer parameters compared to canonical SSMs. When benchmarked against HiPPO-initialized models such as S4 and S4D, an SSM parameterized by Hankel operators demonstrates improved performance on Long-Range Arena (LRA) tasks. Moreover, our new parameterization endows the SSM with non-decaying memory within a fixed time window, which is empirically corroborated by a sequential CIFAR-10 task with padded noise.
title HOPE for a Robust Parameterization of Long-memory State Space Models
topic Machine Learning
url https://arxiv.org/abs/2405.13975