Research Focus

We believe that the solution to the problem of Artificial General Intelligence will be neuro-inspired.

Some representative papers:

Transactions on Machine Learning Research (TMLR), 2025
How Can Knowledge of a Task’s Modular Structure Improve Generalization and Training Efficiency? with Shreyas Malakarjun Patil and Cameron Ethan Taylor.

Transactions on Machine Learning Research (TMLR), 2025
Before Forgetting, There's Learning: Representation Learning Challenges in Online Unsupervised Continual Learning with Cameron Taylor and Shreyas Malakarjun Patil.

IEEE Computer Vision and Pattern Recognition (CVPR) conference, 2024
NICE: Neurogenesis Inspired Contextual Encoding for Replay-free Class Incremental Learning with Mustafa Burak Gurbuz and Jean Michael Mooreman.

Conference on Lifelong Learning Agents (CoLLAs), 2024
Patch-Based Contrastive Learning and Memory Consolidation for Online Unsupervised Continual Learning with Cameron Taylor and Vassilis Vassiliades.

International Conference on Neural Information Processing Systems (NeurIPS) 2023:
Neural Sculpting: Uncovering hierarchically modular task structure in neural networks through pruning and network analysis with Shreyas Malakarjun Patil and Loizos Michael.

International Conference on Machine Learning (ICML) 2022:
NISPA: Neuro-Inspired Stability-Plasticity Adaptation for Continual Learning in Sparse Networks with Mustafa Burak Gurbuz. Code repo: Github link.

International Conference on Machine Learning (ICML) 2021:
PHEW: Constructing Sparse Networks that Learn Fast and Generalize Well without Training Data with Shreyas Malakarjun Patil. Code repo: Github link.

Education

• Ph.D., University of Wisconsin-Madison December 2000
  Research advisor: Parmesh Ramanathan
• M.S., University of Rochester May 1996
Research advisor: Eby Friedman
• B.Eng., Technical University of Crete June 1995
Research advisor: Apostolos Dollas

Teaching (recently)

• CS 7280: Network Science: Methods and Applications
• CS 3510: Design and Analysis of Algorithms (Spring'20). Playlist of YouTube videos from Spring 2021.

Advisees

• Qihang Yao (graduated in Dec'25, PhD CS program, now at Meta)
• Mustafa Burak Gürbüz (graduated in May'25, PhD ML program, now at Meta)
• Cameron Taylor (graduated in May'25, PhD ML program, now at Talon.io)
• Shreyas Malakarjun Patil (graduated in Dec'24, PhD ML program, now at Amazon)
• Kamal Shadi (graduated in Dec'19, now at Decooda)
• Kaeser Md. Sabrin (graduated in Dec'18, now at LinkedIn)
• Payam Siyari (graduated in Dec'18, now at Uber)
• Faryad Sahneh (Postdoc in 2016-17, now at Sanofi)
• Ilias Foudalis (graduated in May'16, now at Relational AI)
• Saamer Akhshabi (Deceased in March 2014. Farewell to Saamer)
• Aemen Lodhi (graduated in Dec'14, now at Facebook)
• Demetris Antoniades (PostDoc 2012-14, now at Cyric)
• Partha Kanuparthy (graduated in Dec'12, now at Amazon)
• Sajjad Zarifzadeh (PostDoc 2010-11, now at National University of Singapore)
• Amogh Dhamdhere (graduated in May'09, now at the CAIDA research center)
• Amund Kvalbein (Postdoc in 2007-08, now at Mason Analysys, Norway)

• Ravi Prasad (graduated in December'07, now at Cisco Systems)
• Ruomei Gao (co-advised with Ellen Zegura, graduated in August'07, now at Akamai)
• Manish Jain (graduated in December'06, now at Akamai)
• Yong Zhu (co-advised with Mostafa Ammar, graduated in August'06, now at Google)
• Qi He (co-advised with Mostafa Ammar, graduated in August'05, now at Apple)

Funding and sponsors (before moving to Cyprus in 2023)

• EAGER:Using Network Analysis And Representational Geometry To Learn Structure-Function Relationship In Neural Networks, Award Number:2039741. Principal Investigators: Eva Dyer and Constantine Dovrolis. Organization: Georgia Tech. NSF Organization:IIS. Start Date:01/01/2021. Award Amount:$199,999.00.

• National Science Foundation,
  "CRCNS: Modeling and Manipulating Dynamic Network Activity in the Brain",
  September 2018 to August 2021. In collaboration with Prof. Shella Keilholz from Emory University.

• DARPA,
  Lifelong Learning Machines (L2M),
  January 2018 to December 2019. In collaboration with Dr. Zsolt Kira (Georgia Tech), Prof. Astrid Prinz (Emory) and Prof. Sarah Pallas (Georgia State Univ).

Open-Source Code (recently)

• PHEW repo. Related paper: PHEW : Constructing Sparse Networks that Learn Fast and Generalize Well without Training Data

• STAM repo. Related paper: Unsupervised Progressive Learning and the STAM Architecture

• C.Elegans hourglass repo. Related paper: The hourglass organization of the C. elegans connectome

• Evo-Lexis project. Related paper: Emergence and Evolution of Hierarchical Structure in Complex Systems

• Lexis project. Related paper: Lexis: An Optimization Framework for Discovering the Hierarchical Structure of Sequential Data

• Delta-Maps project. Related paper: Delta-Maps: from spatio-temporal data to a weighted and lagged network between functional domains

• GENESIS project. Related paper: GENESIS: An agent-based computational model of Internet Interdomain Network Formation, Traffic Flow and Economics