About Me
I am a Ph.D. candidate in Computer Science (CS) working on Machine Learning) at the Free University of Berlin under the supervision of Prof. Eirini Ntoutsi, and also a member of the L3S Research Center in Hannover, Germany. My Ph.D. is a part of the NoBIAS project, a European Innovative Training Network (ITN), funded under the Marie Skłodowska-Curie grant by the European Commission. My research interests lie in the following topics:
- Graph-based learning, Graph Theory, Clustering and Social Networks
- (Un/Semi)supervised learning
- Trustworthy ML: bias and fairness, interpretability, and privacy
- Multi-objective optimization: especially fairness-accuracy trade-off
- Recommendation systems: multi-stakeholder fairness, trade-offs
- Data imbalance: skewness toward a specific class or group
- Representation learning
- Distribution shift, data-stream mining, dynamic graph partitioning
- Address Room 509, Lange Laube 6, 30159 Hannover-Germany
- Email ghodsi@l3s.de
- Phone +49 511 762 16147
- Social Media Google Scholar
- Github
- ORCID
Research Interests
Machine Learning
Data Mining
Multi-objective Optimization
Representation Learning
Fairness-aware Learning
- Data augmentation/generation for bias mitigation
- Fairness-accuracy conflicts and multi-objective trade-offs
- Fair representation learning
- Effect of distribution shifts on bias
Award and Funding
Marie Skłodowska-Curie Innovative Training Network (ITN) H2020 grant
Governmental full tuition-waving fellowship (M.Sc. degree) at University of Kurdistan, Sanadaj-Iran. The fellowship was awarded based on ranking among top 1% students in national university entrance exam.
Governmental full tuition-waving fellowship (B.Sc. degree) at University of Tabriz, Tabriz-Iran. The fellowship was awarded based on ranking among top 1% students in national university entrance exam.
Education
Ph.D. in Computer Science
April 2021 - Present| L3S Research Center (Hannover), & Dept. of Mathematics and Informatics (Free university of Berlin) Hannover - Berlin | Germany Supervisor: Prof. Eirini Ntoutsi
M.Sc. in Artificial Intelligence (AI)
2016 - 2019| Dept. of Computer Engineering, University of Kurdistan, Sanadaj | Iran
Thesis title: Many-objective evolutionary optimization approach using an efficient selection strategy
Supervisor: Prof. Dr. Parham MoradiB.Sc. in Information Technology (IT)
2011 - 2016| Dept. of Electrical and Computer Engineering, University of Tabriz, Tabriz | Iran
Thesis title: Design and implementation of an online bookstore
Supervisor:Community Service
Journal & Conference peer review
2020 | Journal of Information Sciences vol 550-6002020 | Journal of Knowledge-Based Systems special issue on MOA
2023 | 37th Neural Information Processing Systems ethics review (NIPS)
2023 | 26th European Conference on Artificial Intelligence (ECAI 2023)
2024 | ACM Transactions on Knowledge Discovery from Data (TKDD)
Publications
| Authors | Title | Venue | Year | Link |
|---|---|---|---|---|
| Siamak Ghodsi, Seyed Amjad Seyedi, Eirini Ntoutsi | Towards Cohesion-Fairness Harmony: Contrastive Regularization in Individual Fair Graph Clustering | The 28th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD 2024) | 2024 |
|
| Siamak Ghodsi, Eirini Ntoutsi | Affinity Clustering Framework for Data Debiasing Using Pairwise Distribution Discrepancy | Proceedings of the 2nd European Workshop on Algorithmic Fairness (EWAF 2023) | 2023 |
|
| Siamak Ghodsi, Harith Alani, Eirini Ntoutsi | Context matters for fairness - a case study on the effect of spatial distribution shifts | Pre-print (Arxiv) | 2022 |
|
| Seyed Amjad Seyedi, S Siamak Ghodsi, Fardin Akhlaghian, Mahdi Jalili, Parham Moradi | Self-paced multi-label learning with diversity | The 11th Asian Conference on Machine Learning (ACML 2019) | 2019 |
|
| Sahar Tahmasebi, Parham Moradi, Siamak Ghodsi, Alireza Abdollahpouri | An ideal point based many-objective optimization for community detection of complex networks | Journal of Information Sciences | 2019 |
|
| Siamak Ghodsi, Parham Moradi, Sahar Tahmasebi | Graph-based Clustering for Multi-objective Evolutionary Algorithm | 9th International Biannual Symposium on Telecommunications (IST 2018) | 2018 |
|
| Siamak Ghodsi | Poster | NoBIAS Research retreat, Hannover-Germany 14-15 July 2022 | 2022 |
|
News
Our NoBias-ITN paper entitled "Policy advice and best practices on bias and fairness in AI" has been accepted to appear in the Ethics and Information Technology journal (ETIN)
I will be attending the PAKDD 2024 conference to present my paper iFairNMTF