Federated Unlearning with Gradient Descent and Conflict Mitigation Article Swipe
Zibin Pan
,
Zhichao Wang
,
Chi Li
,
Kaiyan Zheng
,
Boqi Wang
,
Xiaoying Tang
,
Junhua Zhao
·
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.1609/aaai.v39i19.34181
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.1609/aaai.v39i19.34181
Federated Learning (FL) has received much attention in recent years. However, although clients are not required to share their data in FL, the global model itself can implicitly remember clients' local data. Therefore, it’s necessary to effectively remove the target client's data from the FL global model to ease the risk of privacy leakage and implement "the right to be forgotten". Federated Unlearning (FU) has been considered a promising solution to remove data without full retraining. But the model utility easily suffers significant reduction during unlearning due to the gradient conflicts. Furthermore, when conducting the post-training to recovery the model utility, it’s prone to move back and revert what have already been unlearned. To address these issues, we propose Federated Unlearning with Orthogonal Steepest Descent (FedOSD). We first design an unlearning cross entropy loss to overcome the convergence issue of the gradient ascent. A steepest descent direction for unlearning is then calculated in the condition of being non-conflicting with other clients’ gradients and closest to the target client's gradient. This benefits to efficiently unlearn and mitigate the model utility reduction. After unlearning, we recover the model utility by maintaining the achievement of unlearning. Finally, extensive experiments in several FL scenarios verify that FedOSD outperforms the SOTA FU algorithms in terms of unlearning and the model utility.
Related Topics
Concepts
Metadata
- Type
- article
- Language
- en
- Landing Page
- https://doi.org/10.1609/aaai.v39i19.34181
- https://ojs.aaai.org/index.php/AAAI/article/download/34181/36336
- OA Status
- diamond
- Cited By
- 4
- Related Works
- 10
- OpenAlex ID
- https://openalex.org/W4409363310
All OpenAlex metadata
Raw OpenAlex JSON
- OpenAlex ID
-
https://openalex.org/W4409363310Canonical identifier for this work in OpenAlex
- DOI
-
https://doi.org/10.1609/aaai.v39i19.34181Digital Object Identifier
- Title
-
Federated Unlearning with Gradient Descent and Conflict MitigationWork title
- Type
-
articleOpenAlex work type
- Language
-
enPrimary language
- Publication year
-
2025Year of publication
- Publication date
-
2025-04-11Full publication date if available
- Authors
-
Zibin Pan, Zhichao Wang, Chi Li, Kaiyan Zheng, Boqi Wang, Xiaoying Tang, Junhua ZhaoList of authors in order
- Landing page
-
https://doi.org/10.1609/aaai.v39i19.34181Publisher landing page
- PDF URL
-
https://ojs.aaai.org/index.php/AAAI/article/download/34181/36336Direct link to full text PDF
- Open access
-
YesWhether a free full text is available
- OA status
-
diamondOpen access status per OpenAlex
- OA URL
-
https://ojs.aaai.org/index.php/AAAI/article/download/34181/36336Direct OA link when available
- Concepts
-
Descent (aeronautics), Computer science, Political science, Environmental science, Geography, MeteorologyTop concepts (fields/topics) attached by OpenAlex
- Cited by
-
4Total citation count in OpenAlex
- Citations by year (recent)
-
2025: 4Per-year citation counts (last 5 years)
- Related works (count)
-
10Other works algorithmically related by OpenAlex
Full payload
| id | https://openalex.org/W4409363310 |
|---|---|
| doi | https://doi.org/10.1609/aaai.v39i19.34181 |
| ids.doi | https://doi.org/10.1609/aaai.v39i19.34181 |
| ids.openalex | https://openalex.org/W4409363310 |
| fwci | 12.86164618 |
| type | article |
| title | Federated Unlearning with Gradient Descent and Conflict Mitigation |
| biblio.issue | 19 |
| biblio.volume | 39 |
| biblio.last_page | 19812 |
| biblio.first_page | 19804 |
| topics[0].id | https://openalex.org/T11612 |
| topics[0].field.id | https://openalex.org/fields/17 |
| topics[0].field.display_name | Computer Science |
| topics[0].score | 0.5630000233650208 |
| topics[0].domain.id | https://openalex.org/domains/3 |
| topics[0].domain.display_name | Physical Sciences |
| topics[0].subfield.id | https://openalex.org/subfields/1702 |
| topics[0].subfield.display_name | Artificial Intelligence |
| topics[0].display_name | Stochastic Gradient Optimization Techniques |
| topics[1].id | https://openalex.org/T12288 |
| topics[1].field.id | https://openalex.org/fields/17 |
| topics[1].field.display_name | Computer Science |
| topics[1].score | 0.5314000248908997 |
| topics[1].domain.id | https://openalex.org/domains/3 |
| topics[1].domain.display_name | Physical Sciences |
| topics[1].subfield.id | https://openalex.org/subfields/1705 |
| topics[1].subfield.display_name | Computer Networks and Communications |
| topics[1].display_name | Optimization and Search Problems |
| topics[2].id | https://openalex.org/T12101 |
| topics[2].field.id | https://openalex.org/fields/18 |
| topics[2].field.display_name | Decision Sciences |
| topics[2].score | 0.4959000051021576 |
| 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 | Advanced Bandit Algorithms Research |
| is_xpac | False |
| apc_list | |
| apc_paid | |
| concepts[0].id | https://openalex.org/C2776637919 |
| concepts[0].level | 2 |
| concepts[0].score | 0.725513219833374 |
| concepts[0].wikidata | https://www.wikidata.org/wiki/Q624380 |
| concepts[0].display_name | Descent (aeronautics) |
| concepts[1].id | https://openalex.org/C41008148 |
| concepts[1].level | 0 |
| concepts[1].score | 0.40955331921577454 |
| concepts[1].wikidata | https://www.wikidata.org/wiki/Q21198 |
| concepts[1].display_name | Computer science |
| concepts[2].id | https://openalex.org/C17744445 |
| concepts[2].level | 0 |
| concepts[2].score | 0.34039098024368286 |
| concepts[2].wikidata | https://www.wikidata.org/wiki/Q36442 |
| concepts[2].display_name | Political science |
| concepts[3].id | https://openalex.org/C39432304 |
| concepts[3].level | 0 |
| concepts[3].score | 0.32150954008102417 |
| concepts[3].wikidata | https://www.wikidata.org/wiki/Q188847 |
| concepts[3].display_name | Environmental science |
| concepts[4].id | https://openalex.org/C205649164 |
| concepts[4].level | 0 |
| concepts[4].score | 0.24750059843063354 |
| concepts[4].wikidata | https://www.wikidata.org/wiki/Q1071 |
| concepts[4].display_name | Geography |
| concepts[5].id | https://openalex.org/C153294291 |
| concepts[5].level | 1 |
| concepts[5].score | 0.1025916337966919 |
| concepts[5].wikidata | https://www.wikidata.org/wiki/Q25261 |
| concepts[5].display_name | Meteorology |
| keywords[0].id | https://openalex.org/keywords/descent |
| keywords[0].score | 0.725513219833374 |
| keywords[0].display_name | Descent (aeronautics) |
| keywords[1].id | https://openalex.org/keywords/computer-science |
| keywords[1].score | 0.40955331921577454 |
| keywords[1].display_name | Computer science |
| keywords[2].id | https://openalex.org/keywords/political-science |
| keywords[2].score | 0.34039098024368286 |
| keywords[2].display_name | Political science |
| keywords[3].id | https://openalex.org/keywords/environmental-science |
| keywords[3].score | 0.32150954008102417 |
| keywords[3].display_name | Environmental science |
| keywords[4].id | https://openalex.org/keywords/geography |
| keywords[4].score | 0.24750059843063354 |
| keywords[4].display_name | Geography |
| keywords[5].id | https://openalex.org/keywords/meteorology |
| keywords[5].score | 0.1025916337966919 |
| keywords[5].display_name | Meteorology |
| language | en |
| locations[0].id | doi:10.1609/aaai.v39i19.34181 |
| locations[0].is_oa | True |
| locations[0].source.id | https://openalex.org/S4210191458 |
| locations[0].source.issn | 2159-5399, 2374-3468 |
| locations[0].source.type | conference |
| locations[0].source.is_oa | True |
| locations[0].source.issn_l | 2159-5399 |
| locations[0].source.is_core | False |
| locations[0].source.is_in_doaj | False |
| locations[0].source.display_name | Proceedings of the AAAI Conference on Artificial Intelligence |
| locations[0].source.host_organization | https://openalex.org/P4310320058 |
| locations[0].source.host_organization_name | Association for the Advancement of Artificial Intelligence |
| locations[0].source.host_organization_lineage | https://openalex.org/P4310320058 |
| locations[0].source.host_organization_lineage_names | Association for the Advancement of Artificial Intelligence |
| locations[0].license | |
| locations[0].pdf_url | https://ojs.aaai.org/index.php/AAAI/article/download/34181/36336 |
| locations[0].version | publishedVersion |
| locations[0].raw_type | journal-article |
| locations[0].license_id | |
| locations[0].is_accepted | True |
| locations[0].is_published | True |
| locations[0].raw_source_name | Proceedings of the AAAI Conference on Artificial Intelligence |
| locations[0].landing_page_url | https://doi.org/10.1609/aaai.v39i19.34181 |
| indexed_in | crossref |
| authorships[0].author.id | https://openalex.org/A5027188224 |
| authorships[0].author.orcid | https://orcid.org/0000-0002-9482-446X |
| authorships[0].author.display_name | Zibin Pan |
| authorships[0].author_position | first |
| authorships[0].raw_author_name | Zibin Pan |
| authorships[0].is_corresponding | False |
| authorships[1].author.id | https://openalex.org/A5106698700 |
| authorships[1].author.orcid | https://orcid.org/0000-0001-8075-1784 |
| authorships[1].author.display_name | Zhichao Wang |
| authorships[1].author_position | middle |
| authorships[1].raw_author_name | Zhichao Wang |
| authorships[1].is_corresponding | False |
| authorships[2].author.id | https://openalex.org/A5100351598 |
| authorships[2].author.orcid | https://orcid.org/0000-0002-8424-9721 |
| authorships[2].author.display_name | Chi Li |
| authorships[2].author_position | middle |
| authorships[2].raw_author_name | Chi Li |
| authorships[2].is_corresponding | False |
| authorships[3].author.id | https://openalex.org/A5115728613 |
| authorships[3].author.orcid | |
| authorships[3].author.display_name | Kaiyan Zheng |
| authorships[3].author_position | middle |
| authorships[3].raw_author_name | Kaiyan Zheng |
| authorships[3].is_corresponding | False |
| authorships[4].author.id | https://openalex.org/A5060367824 |
| authorships[4].author.orcid | |
| authorships[4].author.display_name | Boqi Wang |
| authorships[4].author_position | middle |
| authorships[4].raw_author_name | Boqi Wang |
| authorships[4].is_corresponding | False |
| authorships[5].author.id | https://openalex.org/A5003326020 |
| authorships[5].author.orcid | https://orcid.org/0000-0003-3955-1195 |
| authorships[5].author.display_name | Xiaoying Tang |
| authorships[5].author_position | middle |
| authorships[5].raw_author_name | Xiaoying Tang |
| authorships[5].is_corresponding | False |
| authorships[6].author.id | https://openalex.org/A5065487554 |
| authorships[6].author.orcid | https://orcid.org/0000-0001-5446-2655 |
| authorships[6].author.display_name | Junhua Zhao |
| authorships[6].author_position | last |
| authorships[6].raw_author_name | Junhua Zhao |
| authorships[6].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://ojs.aaai.org/index.php/AAAI/article/download/34181/36336 |
| open_access.oa_status | diamond |
| open_access.any_repository_has_fulltext | False |
| created_date | 2025-10-10T00:00:00 |
| display_name | Federated Unlearning with Gradient Descent and Conflict Mitigation |
| has_fulltext | False |
| is_retracted | False |
| updated_date | 2025-11-06T03:46:38.306776 |
| primary_topic.id | https://openalex.org/T11612 |
| primary_topic.field.id | https://openalex.org/fields/17 |
| primary_topic.field.display_name | Computer Science |
| primary_topic.score | 0.5630000233650208 |
| primary_topic.domain.id | https://openalex.org/domains/3 |
| primary_topic.domain.display_name | Physical Sciences |
| primary_topic.subfield.id | https://openalex.org/subfields/1702 |
| primary_topic.subfield.display_name | Artificial Intelligence |
| primary_topic.display_name | Stochastic Gradient Optimization Techniques |
| related_works | https://openalex.org/W4391375266, https://openalex.org/W2899084033, https://openalex.org/W2748952813, https://openalex.org/W2486267010, https://openalex.org/W4298096494, https://openalex.org/W1678820847, https://openalex.org/W2559216629, https://openalex.org/W1553684505, https://openalex.org/W1607707911, https://openalex.org/W2037211941 |
| cited_by_count | 4 |
| counts_by_year[0].year | 2025 |
| counts_by_year[0].cited_by_count | 4 |
| locations_count | 1 |
| best_oa_location.id | doi:10.1609/aaai.v39i19.34181 |
| best_oa_location.is_oa | True |
| best_oa_location.source.id | https://openalex.org/S4210191458 |
| best_oa_location.source.issn | 2159-5399, 2374-3468 |
| best_oa_location.source.type | conference |
| best_oa_location.source.is_oa | True |
| best_oa_location.source.issn_l | 2159-5399 |
| best_oa_location.source.is_core | False |
| best_oa_location.source.is_in_doaj | False |
| best_oa_location.source.display_name | Proceedings of the AAAI Conference on Artificial Intelligence |
| best_oa_location.source.host_organization | https://openalex.org/P4310320058 |
| best_oa_location.source.host_organization_name | Association for the Advancement of Artificial Intelligence |
| best_oa_location.source.host_organization_lineage | https://openalex.org/P4310320058 |
| best_oa_location.source.host_organization_lineage_names | Association for the Advancement of Artificial Intelligence |
| best_oa_location.license | |
| best_oa_location.pdf_url | https://ojs.aaai.org/index.php/AAAI/article/download/34181/36336 |
| best_oa_location.version | publishedVersion |
| best_oa_location.raw_type | journal-article |
| best_oa_location.license_id | |
| best_oa_location.is_accepted | True |
| best_oa_location.is_published | True |
| best_oa_location.raw_source_name | Proceedings of the AAAI Conference on Artificial Intelligence |
| best_oa_location.landing_page_url | https://doi.org/10.1609/aaai.v39i19.34181 |
| primary_location.id | doi:10.1609/aaai.v39i19.34181 |
| primary_location.is_oa | True |
| primary_location.source.id | https://openalex.org/S4210191458 |
| primary_location.source.issn | 2159-5399, 2374-3468 |
| primary_location.source.type | conference |
| primary_location.source.is_oa | True |
| primary_location.source.issn_l | 2159-5399 |
| primary_location.source.is_core | False |
| primary_location.source.is_in_doaj | False |
| primary_location.source.display_name | Proceedings of the AAAI Conference on Artificial Intelligence |
| primary_location.source.host_organization | https://openalex.org/P4310320058 |
| primary_location.source.host_organization_name | Association for the Advancement of Artificial Intelligence |
| primary_location.source.host_organization_lineage | https://openalex.org/P4310320058 |
| primary_location.source.host_organization_lineage_names | Association for the Advancement of Artificial Intelligence |
| primary_location.license | |
| primary_location.pdf_url | https://ojs.aaai.org/index.php/AAAI/article/download/34181/36336 |
| primary_location.version | publishedVersion |
| primary_location.raw_type | journal-article |
| primary_location.license_id | |
| primary_location.is_accepted | True |
| primary_location.is_published | True |
| primary_location.raw_source_name | Proceedings of the AAAI Conference on Artificial Intelligence |
| primary_location.landing_page_url | https://doi.org/10.1609/aaai.v39i19.34181 |
| publication_date | 2025-04-11 |
| publication_year | 2025 |
| referenced_works_count | 0 |
| abstract_inverted_index.A | 143 |
| abstract_inverted_index.a | 67 |
| abstract_inverted_index.FL | 44, 198 |
| abstract_inverted_index.FU | 206 |
| abstract_inverted_index.To | 113 |
| abstract_inverted_index.We | 126 |
| abstract_inverted_index.an | 129 |
| abstract_inverted_index.be | 59 |
| abstract_inverted_index.by | 187 |
| abstract_inverted_index.in | 7, 20, 152, 196, 208 |
| abstract_inverted_index.is | 149 |
| abstract_inverted_index.of | 51, 139, 155, 191, 210 |
| abstract_inverted_index.to | 16, 35, 47, 58, 70, 87, 96, 103, 134, 164, 171 |
| abstract_inverted_index.we | 117, 182 |
| abstract_inverted_index.But | 76 |
| abstract_inverted_index.FL, | 21 |
| abstract_inverted_index.and | 54, 106, 162, 174, 212 |
| abstract_inverted_index.are | 13 |
| abstract_inverted_index.can | 26 |
| abstract_inverted_index.due | 86 |
| abstract_inverted_index.for | 147 |
| abstract_inverted_index.has | 3, 64 |
| abstract_inverted_index.not | 14 |
| abstract_inverted_index.the | 22, 38, 43, 49, 77, 88, 94, 98, 136, 140, 153, 165, 176, 184, 189, 204, 213 |
| abstract_inverted_index."the | 56 |
| abstract_inverted_index.(FL) | 2 |
| abstract_inverted_index.(FU) | 63 |
| abstract_inverted_index.SOTA | 205 |
| abstract_inverted_index.This | 169 |
| abstract_inverted_index.back | 105 |
| abstract_inverted_index.been | 65, 111 |
| abstract_inverted_index.data | 19, 41, 72 |
| abstract_inverted_index.ease | 48 |
| abstract_inverted_index.from | 42 |
| abstract_inverted_index.full | 74 |
| abstract_inverted_index.have | 109 |
| abstract_inverted_index.loss | 133 |
| abstract_inverted_index.move | 104 |
| abstract_inverted_index.much | 5 |
| abstract_inverted_index.risk | 50 |
| abstract_inverted_index.that | 201 |
| abstract_inverted_index.then | 150 |
| abstract_inverted_index.what | 108 |
| abstract_inverted_index.when | 92 |
| abstract_inverted_index.with | 121, 158 |
| abstract_inverted_index.After | 180 |
| abstract_inverted_index.being | 156 |
| abstract_inverted_index.cross | 131 |
| abstract_inverted_index.data. | 31 |
| abstract_inverted_index.first | 127 |
| abstract_inverted_index.issue | 138 |
| abstract_inverted_index.local | 30 |
| abstract_inverted_index.model | 24, 46, 78, 99, 177, 185, 214 |
| abstract_inverted_index.other | 159 |
| abstract_inverted_index.prone | 102 |
| abstract_inverted_index.right | 57 |
| abstract_inverted_index.share | 17 |
| abstract_inverted_index.terms | 209 |
| abstract_inverted_index.their | 18 |
| abstract_inverted_index.these | 115 |
| abstract_inverted_index.FedOSD | 202 |
| abstract_inverted_index.design | 128 |
| abstract_inverted_index.during | 84 |
| abstract_inverted_index.easily | 80 |
| abstract_inverted_index.global | 23, 45 |
| abstract_inverted_index.itself | 25 |
| abstract_inverted_index.it’s | 33, 101 |
| abstract_inverted_index.recent | 8 |
| abstract_inverted_index.remove | 37, 71 |
| abstract_inverted_index.revert | 107 |
| abstract_inverted_index.target | 39, 166 |
| abstract_inverted_index.verify | 200 |
| abstract_inverted_index.years. | 9 |
| abstract_inverted_index.Descent | 124 |
| abstract_inverted_index.address | 114 |
| abstract_inverted_index.already | 110 |
| abstract_inverted_index.ascent. | 142 |
| abstract_inverted_index.clients | 12 |
| abstract_inverted_index.closest | 163 |
| abstract_inverted_index.descent | 145 |
| abstract_inverted_index.entropy | 132 |
| abstract_inverted_index.issues, | 116 |
| abstract_inverted_index.leakage | 53 |
| abstract_inverted_index.privacy | 52 |
| abstract_inverted_index.propose | 118 |
| abstract_inverted_index.recover | 183 |
| abstract_inverted_index.several | 197 |
| abstract_inverted_index.suffers | 81 |
| abstract_inverted_index.unlearn | 173 |
| abstract_inverted_index.utility | 79, 178, 186 |
| abstract_inverted_index.without | 73 |
| abstract_inverted_index.Finally, | 193 |
| abstract_inverted_index.However, | 10 |
| abstract_inverted_index.Learning | 1 |
| abstract_inverted_index.Steepest | 123 |
| abstract_inverted_index.although | 11 |
| abstract_inverted_index.benefits | 170 |
| abstract_inverted_index.client's | 40, 167 |
| abstract_inverted_index.clients' | 29 |
| abstract_inverted_index.gradient | 89, 141 |
| abstract_inverted_index.mitigate | 175 |
| abstract_inverted_index.overcome | 135 |
| abstract_inverted_index.received | 4 |
| abstract_inverted_index.recovery | 97 |
| abstract_inverted_index.remember | 28 |
| abstract_inverted_index.required | 15 |
| abstract_inverted_index.solution | 69 |
| abstract_inverted_index.steepest | 144 |
| abstract_inverted_index.utility, | 100 |
| abstract_inverted_index.utility. | 215 |
| abstract_inverted_index.(FedOSD). | 125 |
| abstract_inverted_index.Federated | 0, 61, 119 |
| abstract_inverted_index.attention | 6 |
| abstract_inverted_index.condition | 154 |
| abstract_inverted_index.direction | 146 |
| abstract_inverted_index.extensive | 194 |
| abstract_inverted_index.gradient. | 168 |
| abstract_inverted_index.gradients | 161 |
| abstract_inverted_index.implement | 55 |
| abstract_inverted_index.necessary | 34 |
| abstract_inverted_index.promising | 68 |
| abstract_inverted_index.reduction | 83 |
| abstract_inverted_index.scenarios | 199 |
| abstract_inverted_index.Orthogonal | 122 |
| abstract_inverted_index.Therefore, | 32 |
| abstract_inverted_index.Unlearning | 62, 120 |
| abstract_inverted_index.algorithms | 207 |
| abstract_inverted_index.calculated | 151 |
| abstract_inverted_index.clients’ | 160 |
| abstract_inverted_index.conducting | 93 |
| abstract_inverted_index.conflicts. | 90 |
| abstract_inverted_index.considered | 66 |
| abstract_inverted_index.implicitly | 27 |
| abstract_inverted_index.reduction. | 179 |
| abstract_inverted_index.unlearned. | 112 |
| abstract_inverted_index.unlearning | 85, 130, 148, 211 |
| abstract_inverted_index.achievement | 190 |
| abstract_inverted_index.convergence | 137 |
| abstract_inverted_index.effectively | 36 |
| abstract_inverted_index.efficiently | 172 |
| abstract_inverted_index.experiments | 195 |
| abstract_inverted_index.forgotten". | 60 |
| abstract_inverted_index.maintaining | 188 |
| abstract_inverted_index.outperforms | 203 |
| abstract_inverted_index.retraining. | 75 |
| abstract_inverted_index.significant | 82 |
| abstract_inverted_index.unlearning, | 181 |
| abstract_inverted_index.unlearning. | 192 |
| abstract_inverted_index.Furthermore, | 91 |
| abstract_inverted_index.post-training | 95 |
| abstract_inverted_index.non-conflicting | 157 |
| cited_by_percentile_year.max | 98 |
| cited_by_percentile_year.min | 97 |
| countries_distinct_count | 0 |
| institutions_distinct_count | 7 |
| citation_normalized_percentile.value | 0.97085427 |
| citation_normalized_percentile.is_in_top_1_percent | True |
| citation_normalized_percentile.is_in_top_10_percent | True |