analiza aiqbench
MambaTS addresses the quadratic complexity limitations of Transformers in long-term time series forecasting by leveraging linear-scan state space models. The architecture introduces variable-aware scanning strategies and optimized temporal blocks to efficiently model global dependencies without relying on self-attention mechanisms. This approach offers a scalable alternative for handling large-scale sequential data where computational efficiency is critical.
MambaTS: Improved Selective State Space Models for Long-term Time Series Forecasting
arXiv:2405.16440v2 Announce Type: replace Abstract: In recent years, Transformers have become the de-facto architecture for long-term time series forecasting (LTSF), yet they face challenges associated with the self-attention mechanism, including quadratic complexity and permutation-invariant bias. This raises an important question: \emph{do we truly need self-attention to model long-range dependencies in LTSF?} To address this, we propose MambaTS, a linear-scan-based framework that models global dependencies across time and variables via structured dependency modeling. Since explicit variable dependency structures are often unknown, we introduce Variable-Aware Scan along Time (VAST), which learns inter-variable relationships during training and determines an optimal scan order via a shortest-path-based decoding strategy during inference. MambaTS employs the latest Mamba model as its backbone. We suggest that the causal convolution in the vanilla Mamba is unnecessary due to the presence of independent variables, leading to the development of the Temporal Mamba Block (TMB). To mitigate model overfitting, we further incorporate a dropout mechanism for selective parameters in TMB. Extensive experiments conducted on eight public datasets demonstrate that MambaTS achieves competitive or state-of-the-art performance on most datasets. Code is available at this repository: \href{https://github.com/XiudingCai/MambaTS-pytorch}{https://github.com/XiudingCai/MambaTS-pytorch}.
Abstract:In recent years, Transformers have become the de-facto architecture for long-term time series forecasting (LTSF), yet they face challenges associated with the self-attention mechanism, including quadratic complexity and permutation-invariant bias. This raises an important question: \emph{do we truly need self-attention to model long-range dependencies in LTSF?} To address this, we propose MambaTS, a linear-scan-based framework that models global dependencies across time and variables via structured dependency modeling. Since explicit variable dependency structures are often unknown, we introduce Variable-Aware Scan along Time (VAST), which learns inter-variable relationships during training and determines an optimal scan order via a shortest-path-based decoding strategy during inference. MambaTS employs the latest Mamba model as its backbone. We suggest that the causal convolution in the vanilla Mamba is unnecessary due to the presence of independent variables, leading to the development of the Temporal Mamba Block (TMB). To mitigate model overfitting, we further incorporate a dropout mechanism for selective parameters in TMB. Extensive experiments conducted on eight public datasets demonstrate that MambaTS achieves competitive or state-of-the-art performance on most datasets. Code is available at this repository: \href{this https URL}{this https URL}.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2405.16440 [cs.LG]
(or arXiv:2405.16440v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2405.16440 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 278 cuvinte
