Amjad Seyedi

Research interests

• Machine Learning: representation learning, deep learning, and unsupervised learning
• Trustworthy ML: robustness, generalization, interpretability, and fairness
• Applications: healthcare, recommender systems, and remote sensing
• Applied Mathematics: linear algebra, optimization, and low-rank approximation

News & updates

• Jan–Mar 2027 · Upcoming — Visiting researcher at Inria Montpellier. Planned research stay in Montpellier, hosted by Aurélien Bellet, focusing on low-rank methods for federated learning.
• Oct 2026 — Presentation at ML4Hyper3D 2026. Presented work on Bilinear Nonnegative Matrix Factorization for Hyperspectral Unmixing in Lyon, France.
• Oct 5, 2026 — Optimization for Interpretable Machine Learning workshop. Attended the workshop on classification, clustering, and interpretable machine learning at ESSEC Business School, Cergy, France.
• Sep 7, 2026 — SOCN Study Day in Namur. Attended the Graduate School in Systems, Optimization, Control and Networks Study Day in Namur, Belgium.
• Aug–Sep 2026 — Presentation at EUSIPCO 2026. Presented Bilinear Nonnegative Matrix Factorization for Hyperspectral Unmixing at EUSIPCO in Bruges, Belgium.
• 2026 — Paper accepted in Expert Systems with Applications. Distributionally Robust Multi-View Clustering with Anchor Graph Learning, scheduled for the 2027 volume.
• Jul 2026 — Attended FoCM 2026. Attended the Foundations of Computational Mathematics meeting in Vienna, Austria (July 16–18).
• Jun 24, 2026 — Completed PhD confirmation. Successfully completed doctoral confirmation in Matrix Theory and Optimization at the University of Mons.
• 2026 — Paper accepted at EUSIPCO 2026. Bilinear Nonnegative Matrix Factorization for Hyperspectral Unmixing accepted for the European Signal Processing Conference.
• May 2026 — Presentation at ICASSP 2026. Presented Encoder-Decoder Symmetric Nonnegative Matrix Tri-Factorization for Graph Clustering in Barcelona, Spain.
• 2026 — Paper accepted in Neurocomputing. Robust Asymmetric Encoder-Decoder Nonnegative Matrix Factorization for Hyperspectral Anomaly Detection.
• 2026 — Paper accepted in Knowledge-Based Systems. Robust log-based multi-label feature selection with dynamic label correlation and relevance–redundancy optimization.
• 2026 — Pattern Recognition paper accepted: Contrastive Calibration. Contrastive Calibration on Consensus and Complementary Multi-View Representations.
• 2026 — Paper accepted at ICASSP 2026. Encoder-Decoder Symmetric Nonnegative Matrix Tri-Factorization for Graph Clustering.
• 2026 — Paper accepted in International Journal of Machine Learning and Cybernetics. Semantic Encoder-Decoder Nonnegative Matrix Factorization with Kullback–Leibler Divergence.
• 2026 — Paper accepted in Engineering Applications of Artificial Intelligence. Community Detection via Deep Motif-regularized Asymmetric Nonnegative Matrix Factorization.
• Oct 2025 — CWI Autumn School in Scientific Machine Learning. Attended Scientific Machine Learning and Numerical Methods at Centrum Wiskunde & Informatica, Netherlands (Oct 27–31).
• 2025 — Paper accepted at IEEE BigData 2025. A Deep Latent Factor Graph Clustering with Fairness-Utility Trade-off Perspective.
• 2025 — Paper accepted in Information Sciences. Distributionally Robust Nonnegative Matrix Factorization with Self-Paced Adaptive Multi-Loss Fusion (2026 volume).
• Sep–Oct 2025 — Three MSc thesis defenses. Three master’s students from the Algebraic Machine Learning team at the University of Kurdistan successfully defended their theses.
• Oct 2025 — SOCN course on stochastic optimization and federated learning. Attended the SOCN course at UCLouvain, Belgium (Oct 7–9).
• Sep 26, 2025 — SOCN workshop at UMONS. Attended the Systems, Optimization, Control and Networks workshop in Mons, Belgium.
• Sep 2025 — Organizing committee of LRMA 25. Served on the organizing committee of the third Low-Rank Models and Applications workshop in Mons, Belgium (Sep 11–12).
• Sep 2025 — CWI PhD School on Machine Learning and Optimization. Attended the CWI PhD School in the Netherlands (Sep 2–4).
• 2025 — Pattern Recognition paper accepted: Encoder-Decoder NMF. Encoder-Decoder Nonnegative Matrix Factorization with β-Divergence for Data Clustering (2026 volume).
• 2025 — Paper accepted in Pattern Recognition: instance-wise DRNMF. Instance-wise Distributionally Robust Nonnegative Matrix Factorization (2026 volume), a collaboration with Prof. Nicolas Gillis.
• Apr 2025 — Learning Theory from First Principles at KU Leuven. Attended the course by Francis Bach at KU Leuven, Belgium (Apr 1–3).
• Mar 2025 — Masterclass on Graphs, Data and AI. Attended the masterclass with Jure Leskovec at the University of Antwerp, Belgium (Mar 20).
• Feb 20, 2025 — Invited talk at STADIUS, KU Leuven. Delivered a talk on Deep Matrix Factorizations for Data Representation.
• Oct 29, 2024 — SOCN Study Day at UCLouvain. Attended the Graduate School in Systems, Optimization, Control and Networks Study Day in Belgium.
• Oct 2024 — Autumn School in Constrained Optimization and Machine Learning. Attended the school at Trier University, Germany (Oct 9–11).
• Aug 2024 — ALGOPT2024 workshop. Attended the workshop on Algorithmic Optimization: Tools for AI and Data Science at UCLouvain, Belgium (Aug 27–30).
• Aug 2024 — Poster presentation accepted at ALGOPT2024. A poster on deep asymmetric nonnegative matrix factorization for graph clustering was accepted for the workshop.
• Jun 18, 2024 — Started PhD at University of Mons. Started doctoral research in Matrix Theory and Optimization at Polytech MONS, Belgium.

Education

• PhD in Mathematics & Operational Research — University of Mons (UMONS), Mons, Belgium (Jun 2024 – Present). Thesis: Advancing Matrix Factorization Models for Trustworthy Machine Learning and Signal Processing · Advisor: Prof. Nicolas Gillis | Co-advisor: Prof. Arnaud Vandaele · PhD confirmation: June 2026 | Expected defense: June 2027 · Topics: low-rank approximation, optimization, robust and trustworthy machine learning
• Master's Degree in Artificial Intelligence — University of Kurdistan, Sanandaj, Iran (Sep 2015 – Feb 2018). Thesis: "A Graph-based Semi-Supervised Learning Approach for Multi-Label Classification" · Supervisor: Dr. Parham Moradi | Advisor: Dr. Fardin Akhlaghian Tab · Topics: semi-supervised learning, multi-label learning, matrix factorization
• Bachelor's Degree in Software Engineering — Amirkabir Technical College, Arak, Iran (Jan 2012 – Jun 2014). Project: Manufacturing and Setting up a Video Conferencing Software
• Associate Degree in Computer Software — Technical College of Tabriz, Tabriz, Iran (Jan 2009 – Jun 2011).

Experience

• Visiting Researcher (scheduled) — Inria, Montpellier, France (Jan 2027 – Mar 2027). Host: Dr. Aurélien Bellet · Topic: Low-rank matrix factorization methods for federated learning.
• Team Leader & Thesis Advisor — Algebraic Machine Learning (AML) Team, University of Kurdistan, Iran (Sep 2020 – Jun 2024). Led research in representation learning, matrix factorization, semi-supervised learning, robustness and sparse coding. Supported 13 completed master’s theses; currently advising one PhD and one master’s student.
• Graduate Research Assistant — Representation Learning Lab, University of Kurdistan, Iran (Sep 2019 – Jun 2024). Research on matrix factorization, distributionally robust learning, generalization and adversarial training, including image inpainting and recommender systems.
• Teaching Assistant — University of Kurdistan, Iran (Jan 2019 – Jun 2024). Advanced Concepts in Artificial Intelligence (Graduate), Spring & Fall 2023 · Nonnegative Matrix Factorization for Machine Learning (Graduate), Fall 2022 · Pattern Recognition (Graduate), Spring 2019–Spring 2023 · Special Topics in Artificial Intelligence (Graduate), Fall 2021
• Lab Instructor — University of Kurdistan, Iran (Fall 2019). Course: Computer Lab

Students & research mentoring

• Wafa Barkhoda — PhD during mentorship (2022–Present); Distributionally robust optimization; current position: Faculty member at Kermanshah University of Technology
• Zahra Mohsenirad — MSc during mentorship (2023–2026); Hyperspectral unmixing
• Shirin Moradi — MSc during mentorship (2022–2025); Anomaly detection in hyperspectral images
• Sarina Khaledian — MSc during mentorship (2022–2025); Popularity bias, diversity, and fairness in recommendation
• Negin Jabari — MSc during mentorship (2023–2025); Multi-view representation
• Sayvan Soleymanbeigi — MSc during mentorship (2022–2024); Topic modeling, text clustering; current position: Generative AI Engineer at Trifay Hataw
• Arina Mohammadi — MSc during mentorship (2021–2023); Attributed graph clustering
• Mohammad Faraji — MSc during mentorship (2021–2023); Multi-label feature selection
• Akram Hajiveiseh — MSc during mentorship (2020–2023); Directed graph clustering; current position: Research fellow at Edinburgh Napier University
• Maryam Mozafari — MSc during mentorship (2021–2022); Unsupervised feature selection; current position: PhD student at University of Quebec in Outaouais

Teaching

• Advanced Concepts in Artificial Intelligence (Graduate) — Spring and Fall 2023
• Nonnegative Matrix Factorization for Machine Learning (Graduate) — Fall 2022
• Pattern Recognition (Graduate) — Spring 2019 – Spring 2023
• Special Topics in Artificial Intelligence (Graduate) — Fall 2021
• Computer Lab (Undergraduate) — Fall 2019 — two sections

Talks and presentations

• Oct 2026 — ML4Hyper3D 2026, Lyon, France: Bilinear Nonnegative Matrix Factorization for Hyperspectral Unmixing
• Aug–Sep 2026 — EUSIPCO 2026, Bruges, Belgium: Bilinear Nonnegative Matrix Factorization for Hyperspectral Unmixing
• May 2026 — ICASSP 2026, Barcelona, Spain: Encoder-Decoder Symmetric Nonnegative Matrix Tri-Factorization for Graph Clustering
• Sep 2025 — LRMA 2025, Mons, Belgium: Instance-wise distributionally robust nonnegative matrix factorization
• Feb 2025 — Invited talk, STADIUS, KU Leuven, Belgium: Deep Matrix Factorizations for Data Representation
• Aug 2024 — ALGOPT2024, UCLouvain, Belgium: Deep asymmetric nonnegative matrix factorization for graph clustering

Courses and schools

• Oct 2025: CWI Autumn School: Scientific Machine Learning and Numerical Methods — CWI, Netherlands
• Oct 2025: SOCN: Stochastic Variance-Reduced Optimization Algorithms and Applications to Federated Learning — UCLouvain, Belgium
• Sep 2025: CWI PhD School: Machine Learning and Optimization — CWI, Netherlands
• Apr 2025: SOCN: Learning Theory from First Principles (Francis Bach) — KU Leuven, Belgium
• Mar 2025: Masterclass: Graphs, Data and AI (Jure Leskovec) — University of Antwerp, Belgium
• Oct 2024: Autumn School: Constrained Optimization and Machine Learning — Trier University, Germany

Journal reviewing

• Artificial Intelligence Review
• Computers and Electrical Engineering
• Digital Signal Processing
• Engineering Applications of Artificial Intelligence
• Expert Systems with Applications
• Information Fusion
• Information Processing & Management
• Information Sciences
• International Journal of Machine Learning and Cybernetics
• Journal of Computational and Applied Mathematics
• Journal of Information and Intelligence
• Knowledge-Based Systems
• Mathematics
• Medical Image Analysis
• Neurocomputing
• IEEE Transactions on Neural Networks and Learning Systems
• Pattern Recognition
• Scientific Reports
• Signal Processing
• ACM Transactions on Intelligent Systems and Technology

Conference reviewing

• AAAI Conference on Artificial Intelligence (AAAI), 2027 — 6 papers
• IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2027 — 4 papers
• International Conference on Machine Learning (ICML), 2026 — 5 papers, Silver Reviewer
• IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2026

Technical skills

• Python, PyTorch, NumPy, scikit-learn
• MATLAB
• C++, C#
• LaTeX, Manim
• Jupyter, Colab, PyCharm
• Windows, Linux
• Adobe Illustrator, Inkscape, Photoshop, GIMP

Publications and preprints (31)

1. Supervised Deep Multimodal Matrix Factorization for Interpretable Brain Network Analysis
   A. Seyedi , Lifang He, Songlin Zhao, Akwum Onwunta, and Nicolas Gillis.
2. ECG-NAT: A Self-supervised Neighborhood Attention Transformer for Multi-lead ECG Classification
   M. Gazeran, S. Soleymanbaigi, F. Daneshfar, A. Seyedi , and F. Akhlaghian.
3. Distributionally Robust Multi-View Clustering with Anchor Graph Learning
   W. Barkhoda, A. Seyedi , and F. Akhlaghian, Expert Systems with Applications 2027 .
4. Bilinear Nonnegative Matrix Factorization for Hyperspectral Unmixing
   A. Seyedi , A. Vandaele, and N. Gillis, EUSIPCO 2026 .
   Source Code
5. Robust Asymmetric Encoder-Decoder Nonnegative Matrix Factorization for Hyperspectral Anomaly Detection
   S. Moradi, A. Seyedi , W. Barkhoda, and F. Akhlaghian, Neurocomputing 2026 .
   Source Code
6. Robust log-based multi-label feature selection with dynamic label correlation and relevance–redundancy optimization
   M. Faraji, A. Seyedi , and F. Akhlaghian, Knowledge-Based Systems 2026 .
   Source Code
7. Contrastive Calibration on Consensus and Complementary Multi-View Representations
   N. Jabari, A. Seyedi , R. Mahmoodi, and F. Akhlaghian, Pattern Recognition 2026 .
   Source Code
8. Encoder-Decoder Symmetric Nonnegative Matrix Tri-Factorization for Graph Clustering
   A. Seyedi and N. Gillis, ICASSP 2026 .
   Source Code
9. Semantic Encoder-Decoder Nonnegative Matrix Factorization with Kullback-Leibler Divergence
   S. Soleymanbaigi, A. Seyedi , F. Daneshfar, and F. Akhlaghian, IJMLC 2026 .
10. Community Detection via Deep Motif-regularized Asymmetric Nonnegative Matrix Factorization
    H. Sohrabi, A. Seyedi , P. Moradi, and S. Esmaeili, Engineering Applications of AI 2026 .
    Source Code
11. Distributionally Robust Nonnegative Matrix Factorization with Self-Paced Adaptive Multi-Loss Fusion
    W. Barkhoda, A. Seyedi , N. Gillis, and F. Akhlaghian, Information Sciences 2026 .
    Source Code
12. Encoder-Decoder Nonnegative Matrix Factorization with β-Divergence for Data Clustering
    S. Soleymanbaigi, A. Seyedi , F. Akhlaghian, and F. Daneshfar, Pattern Recognition 2026 .
13. Instance-wise Distributionally Robust Nonnegative Matrix Factorization
    W. Barkhoda, A. Seyedi , N. Gillis, and F. Akhlaghian, Pattern Recognition 2026 .
    Source Code
14. A Deep Latent Factor Graph Clustering with Fairness-Utility Trade-off Perspective
    S. Ghodsi, A. Seyedi , T. Quy, F. Karimi, and E. Ntoutsi, IEEE BigData 2025 .
    Source Code
15. A New Bi-Level Deep Human Action Representation Structure based on the Sequence of Sub-Actions
    F. Akhlaghian, M. Ramezani, H. Afshoon, A. Seyedi , and A. Moradyani, Neural Computing and Applications 2025 .
16. Diverse Joint Nonnegative Matrix Factorization for Attributed Graph Clustering
    A. Mohammadi, A. Seyedi , F. Akhlaghian, and R. Pir Mohammadiani, Applied Soft Computing 2024 .
17. Enhancing Link Prediction through Adversarial Training in Deep Nonnegative Matrix Factorization
    R. Mahmoodi, A. Seyedi , A. Abdollahpouri, and F. Akhlaghian, Engineering Applications of AI 2024 .
18. Towards Cohesion-Fairness Harmony: Contrastive Regularization in Individual Fair Graph Clustering
    S. Ghodsi, A. Seyedi , and E. Ntoutsi, PAKDD 2024 .
    Source Code · PDF
19. Orthogonal Encoder-Decoder Factorization for Unsupervised Feature Selection
    M. Mozafari, A. Seyedi , R. Pir Mohammadiani, and F. Akhlaghian, Information Sciences 2024 .
20. Multi-Label Feature Selection with Global and Local Label Correlation
    M. Faraji, A. Seyedi , F. Akhlaghian, and R. Mahmoodi, Expert Systems with Applications 2024 .
21. Deep Asymmetric Nonnegative Matrix Factorization for Graph Clustering
    A. Hajiveiseh, A. Seyedi , and F. Akhlaghian, Pattern Recognition 2024 .
22. Link Prediction by Adversarial Nonnegative Matrix Factorization
    R. Mahmoodi, A. Seyedi , F. Akhlaghian, and A. Abdollahpouri, Knowledge-Based Systems 2023 .
23. Self-Supervised Semi-Supervised Nonnegative Matrix Factorization for Data Clustering
    Jovan Chavoshinejad, Seyed Amjad Seyedi , Fardin Akhlaghian Tab, and Navid Salahian, Pattern Recognition 2023 .
    Source Code
24. Adversarial Elastic Deep Nonnegative Matrix Factorization for Matrix Completion
    Seyed Amjad Seyedi , Fardin Akhlaghian Tab, Abdulrahman Lotfi, Navid Salahian, and Jovan Chavoshinejad, Information Sciences 2023 .
    Source Code
25. Deep Autoencoder-Like NMF with Contrastive Regularization and Feature Relationship Preservation
    Navid Salahian, Fardin Akhlaghian Tab, Seyed Amjad Seyedi , and Jovan Chavoshinejad, Expert Systems with Applications 2023 .
26. Asymmetric Semi-Nonnegative Matrix Factorization for Directed Graph Clustering
    R. Abdollahi, A. Seyedi , and M. R. Noorimehr, ICCKE 2020 .
27. A Weakly-Supervised Factorization Method with Dynamic Graph Embedding
    A. Seyedi , P. Moradi, and F. Akhlaghian Tab, AISP 2017 .
28. A Clustering-based Matrix Factorization Method to Improve the Accuracy of Recommendation Systems
    Z. Shajarian, A. Seyedi , and P. Moradi, ICEE 2017 .
29. An Improved Density Peaks Method for Data Clustering
    A. Lotfi, A. Seyedi , and P. Moradi, ICCKE 2016 .
30. Self-Paced Multi-Label Learning with Diversity
    A. Seyedi , S. Ghodsi, F. Akhlaghian, M. Jalili, and P. Moradi, ACML 2019 .
    Source Code
31. Dynamic Graph-based Label Propagation for Density Peaks Clustering
    A. Seyedi , A. Lotfi, P. Moradi, and N.N. Qader, Expert Systems with Applications 2019 .

Academic references

• Prof. Nicolas Gillis — PhD supervisor, University of Mons · nicolas.gillis@umons.ac.be
• Prof. Arnaud Vandaele — PhD co-advisor, University of Mons · arnaud.vandaele@umons.ac.be
• Dr. Fardin Akhlaghian — MSc advisor and collaborator, University of Kurdistan · f.akhlaghian@uok.ac.ir
• Dr. Parham Moradi — MSc supervisor, Victoria University · parham.moradi@vu.edu.au