Mixture of Low Rank Adaptation with Partial Parameter Sharing for Time Series Forecasting Article Swipe

PDF

Linfeng Pan , Zhichao Chen , Haoxuan Li , Guangyi Liu , Zhijian Xu , Zhiqiang Liu , Hao Wang , Ying Wei ·

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2505.17872

Multi-task forecasting has become the standard approach for time-series forecasting (TSF). However, we show that it suffers from an Expressiveness Bottleneck, where predictions at different time steps share the same representation, leading to unavoidable errors even with optimal representations. To address this issue, we propose a two-stage framework: first, pre-train a foundation model for one-step-ahead prediction; then, adapt it using step-specific LoRA modules.This design enables the foundation model to handle any number of forecast steps while avoiding the expressiveness bottleneck. We further introduce the Mixture-of-LoRA (MoLA) model, which employs adaptively weighted LoRA experts to achieve partial parameter sharing across steps. This approach enhances both efficiency and forecasting performance by exploiting interdependencies between forecast steps. Experiments show that MoLA significantly improves model expressiveness and outperforms state-of-the-art time-series forecasting methods. Code is available at https://anonymous.4open.science/r/MoLA-BC92.

Related Topics

Concepts

No concepts available.

Metadata

Type: preprint
Language: en
Landing Page: http://arxiv.org/abs/2505.17872
PDF: https://arxiv.org/pdf/2505.17872
OA Status: green
OpenAlex ID: https://openalex.org/W4414580819

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4414580819

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2505.17872

Digital Object Identifier
Title: Mixture of Low Rank Adaptation with Partial Parameter Sharing for Time Series Forecasting

Work title
Type: preprint

OpenAlex work type
Language: en

Primary language
Publication year: 2025

Year of publication
Publication date: 2025-05-23

Full publication date if available
Authors: Linfeng Pan, Zhichao Chen, Haoxuan Li, Guangyi Liu, Zhijian Xu, Zhiqiang Liu, Hao Wang, Ying Wei

List of authors in order
Landing page: https://arxiv.org/abs/2505.17872

Publisher landing page
PDF URL: https://arxiv.org/pdf/2505.17872

Direct link to full text PDF
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2505.17872

Direct OA link when available
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W4414580819
doi	https://doi.org/10.48550/arxiv.2505.17872
ids.doi	https://doi.org/10.48550/arxiv.2505.17872
ids.openalex	https://openalex.org/W4414580819
fwci
type	preprint
title	Mixture of Low Rank Adaptation with Partial Parameter Sharing for Time Series Forecasting
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
topics[0].id	https://openalex.org/T12205
topics[0].field.id	https://openalex.org/fields/17
topics[0].field.display_name	Computer Science
topics[0].score	0.9451000094413757
topics[0].domain.id	https://openalex.org/domains/3
topics[0].domain.display_name	Physical Sciences
topics[0].subfield.id	https://openalex.org/subfields/1711
topics[0].subfield.display_name	Signal Processing
topics[0].display_name	Time Series Analysis and Forecasting
topics[1].id	https://openalex.org/T11918
topics[1].field.id	https://openalex.org/fields/18
topics[1].field.display_name	Decision Sciences
topics[1].score	0.9090999960899353
topics[1].domain.id	https://openalex.org/domains/2
topics[1].domain.display_name	Social Sciences
topics[1].subfield.id	https://openalex.org/subfields/1803
topics[1].subfield.display_name	Management Science and Operations Research
topics[1].display_name	Forecasting Techniques and Applications
topics[2].id	https://openalex.org/T11326
topics[2].field.id	https://openalex.org/fields/18
topics[2].field.display_name	Decision Sciences
topics[2].score	0.9072999954223633
topics[2].domain.id	https://openalex.org/domains/2
topics[2].domain.display_name	Social Sciences
topics[2].subfield.id	https://openalex.org/subfields/1803
topics[2].subfield.display_name	Management Science and Operations Research
topics[2].display_name	Stock Market Forecasting Methods
is_xpac	False
apc_list
apc_paid
language	en
locations[0].id	pmh:oai:arXiv.org:2505.17872
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S4306400194
locations[0].source.issn
locations[0].source.type	repository
locations[0].source.is_oa	True
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	arXiv (Cornell University)
locations[0].source.host_organization	https://openalex.org/I205783295
locations[0].source.host_organization_name	Cornell University
locations[0].source.host_organization_lineage	https://openalex.org/I205783295
locations[0].license	cc-by
locations[0].pdf_url	https://arxiv.org/pdf/2505.17872
locations[0].version	submittedVersion
locations[0].raw_type	text
locations[0].license_id	https://openalex.org/licenses/cc-by
locations[0].is_accepted	False
locations[0].is_published	False
locations[0].raw_source_name
locations[0].landing_page_url	http://arxiv.org/abs/2505.17872
locations[1].id	doi:10.48550/arxiv.2505.17872
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published
locations[1].raw_source_name
locations[1].landing_page_url	https://doi.org/10.48550/arxiv.2505.17872
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5102995307
authorships[0].author.orcid	https://orcid.org/0000-0003-0398-7333
authorships[0].author.display_name	Linfeng Pan
authorships[0].author_position	first
authorships[0].raw_author_name	Pan, Licheng
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5100619476
authorships[1].author.orcid	https://orcid.org/0000-0002-8431-6066
authorships[1].author.display_name	Zhichao Chen
authorships[1].author_position	middle
authorships[1].raw_author_name	Chen, Zhichao
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5100629474
authorships[2].author.orcid	https://orcid.org/0000-0003-3903-7364
authorships[2].author.display_name	Haoxuan Li
authorships[2].author_position	middle
authorships[2].raw_author_name	Li, Haoxuan
authorships[2].is_corresponding	False
authorships[3].author.id	https://openalex.org/A5100737328
authorships[3].author.orcid	https://orcid.org/0000-0001-9822-2039
authorships[3].author.display_name	Guangyi Liu
authorships[3].author_position	middle
authorships[3].raw_author_name	Liu, Guangyi
authorships[3].is_corresponding	False
authorships[4].author.id	https://openalex.org/A5100512419
authorships[4].author.orcid	https://orcid.org/0009-0006-0527-1442
authorships[4].author.display_name	Zhijian Xu
authorships[4].author_position	middle
authorships[4].raw_author_name	Xu, Zhijian
authorships[4].is_corresponding	False
authorships[5].author.id	https://openalex.org/A5100730557
authorships[5].author.orcid	https://orcid.org/0000-0001-8752-5755
authorships[5].author.display_name	Zhiqiang Liu
authorships[5].author_position	middle
authorships[5].raw_author_name	Liu, Zhaoran
authorships[5].is_corresponding	False
authorships[6].author.id	https://openalex.org/A5116592778
authorships[6].author.orcid	https://orcid.org/0000-0003-2227-075X
authorships[6].author.display_name	Hao Wang
authorships[6].author_position	middle
authorships[6].raw_author_name	Wang, Hao
authorships[6].is_corresponding	False
authorships[7].author.id	https://openalex.org/A5100619276
authorships[7].author.orcid	https://orcid.org/0000-0002-4247-1770
authorships[7].author.display_name	Ying Wei
authorships[7].author_position	last
authorships[7].raw_author_name	Wei, Ying
authorships[7].is_corresponding	False
has_content.pdf	True
has_content.grobid_xml	True
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2505.17872
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-10T00:00:00
display_name	Mixture of Low Rank Adaptation with Partial Parameter Sharing for Time Series Forecasting
has_fulltext	False
is_retracted	False
updated_date	2025-11-06T06:51:31.235846
primary_topic.id	https://openalex.org/T12205
primary_topic.field.id	https://openalex.org/fields/17
primary_topic.field.display_name	Computer Science
primary_topic.score	0.9451000094413757
primary_topic.domain.id	https://openalex.org/domains/3
primary_topic.domain.display_name	Physical Sciences
primary_topic.subfield.id	https://openalex.org/subfields/1711
primary_topic.subfield.display_name	Signal Processing
primary_topic.display_name	Time Series Analysis and Forecasting
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2505.17872
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license	cc-by
best_oa_location.pdf_url	https://arxiv.org/pdf/2505.17872
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id	https://openalex.org/licenses/cc-by
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2505.17872
primary_location.id	pmh:oai:arXiv.org:2505.17872
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S4306400194
primary_location.source.issn
primary_location.source.type	repository
primary_location.source.is_oa	True
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	arXiv (Cornell University)
primary_location.source.host_organization	https://openalex.org/I205783295
primary_location.source.host_organization_name	Cornell University
primary_location.source.host_organization_lineage	https://openalex.org/I205783295
primary_location.license	cc-by
primary_location.pdf_url	https://arxiv.org/pdf/2505.17872
primary_location.version	submittedVersion
primary_location.raw_type	text
primary_location.license_id	https://openalex.org/licenses/cc-by
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	http://arxiv.org/abs/2505.17872
publication_date	2025-05-23
publication_year	2025
referenced_works_count	0
abstract_inverted_index.a	45, 50
abstract_inverted_index.To	39
abstract_inverted_index.We	80
abstract_inverted_index.an	18
abstract_inverted_index.at	23, 131
abstract_inverted_index.by	108
abstract_inverted_index.is	129
abstract_inverted_index.it	15, 58
abstract_inverted_index.of	72
abstract_inverted_index.to	32, 68, 93
abstract_inverted_index.we	12, 43
abstract_inverted_index.and	105, 122
abstract_inverted_index.any	70
abstract_inverted_index.for	7, 53
abstract_inverted_index.has	2
abstract_inverted_index.the	4, 28, 65, 77, 83
abstract_inverted_index.Code	128
abstract_inverted_index.LoRA	61, 91
abstract_inverted_index.MoLA	117
abstract_inverted_index.This	100
abstract_inverted_index.both	103
abstract_inverted_index.even	35
abstract_inverted_index.from	17
abstract_inverted_index.same	29
abstract_inverted_index.show	13, 115
abstract_inverted_index.that	14, 116
abstract_inverted_index.this	41
abstract_inverted_index.time	25
abstract_inverted_index.with	36
abstract_inverted_index.adapt	57
abstract_inverted_index.model	52, 67, 120
abstract_inverted_index.share	27
abstract_inverted_index.steps	26, 74
abstract_inverted_index.then,	56
abstract_inverted_index.using	59
abstract_inverted_index.where	21
abstract_inverted_index.which	87
abstract_inverted_index.while	75
abstract_inverted_index.(MoLA)	85
abstract_inverted_index.(TSF).	10
abstract_inverted_index.across	98
abstract_inverted_index.become	3
abstract_inverted_index.design	63
abstract_inverted_index.errors	34
abstract_inverted_index.first,	48
abstract_inverted_index.handle	69
abstract_inverted_index.issue,	42
abstract_inverted_index.model,	86
abstract_inverted_index.number	71
abstract_inverted_index.steps.	99, 113
abstract_inverted_index.achieve	94
abstract_inverted_index.address	40
abstract_inverted_index.between	111
abstract_inverted_index.employs	88
abstract_inverted_index.enables	64
abstract_inverted_index.experts	92
abstract_inverted_index.further	81
abstract_inverted_index.leading	31
abstract_inverted_index.optimal	37
abstract_inverted_index.partial	95
abstract_inverted_index.propose	44
abstract_inverted_index.sharing	97
abstract_inverted_index.suffers	16
abstract_inverted_index.However,	11
abstract_inverted_index.approach	6, 101
abstract_inverted_index.avoiding	76
abstract_inverted_index.enhances	102
abstract_inverted_index.forecast	73, 112
abstract_inverted_index.improves	119
abstract_inverted_index.methods.	127
abstract_inverted_index.standard	5
abstract_inverted_index.weighted	90
abstract_inverted_index.available	130
abstract_inverted_index.different	24
abstract_inverted_index.introduce	82
abstract_inverted_index.parameter	96
abstract_inverted_index.pre-train	49
abstract_inverted_index.two-stage	46
abstract_inverted_index.Multi-task	0
abstract_inverted_index.adaptively	89
abstract_inverted_index.efficiency	104
abstract_inverted_index.exploiting	109
abstract_inverted_index.foundation	51, 66
abstract_inverted_index.framework:	47
abstract_inverted_index.Bottleneck,	20
abstract_inverted_index.Experiments	114
abstract_inverted_index.bottleneck.	79
abstract_inverted_index.forecasting	1, 9, 106, 126
abstract_inverted_index.outperforms	123
abstract_inverted_index.performance	107
abstract_inverted_index.prediction;	55
abstract_inverted_index.predictions	22
abstract_inverted_index.time-series	8, 125
abstract_inverted_index.unavoidable	33
abstract_inverted_index.modules.This	62
abstract_inverted_index.significantly	118
abstract_inverted_index.step-specific	60
abstract_inverted_index.Expressiveness	19
abstract_inverted_index.expressiveness	78, 121
abstract_inverted_index.one-step-ahead	54
abstract_inverted_index.Mixture-of-LoRA	84
abstract_inverted_index.representation,	30
abstract_inverted_index.representations.	38
abstract_inverted_index.state-of-the-art	124
abstract_inverted_index.interdependencies	110
abstract_inverted_index.https://anonymous.4open.science/r/MoLA-BC92.	132
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	8
citation_normalized_percentile