I hold a PhD in signal processing and telecommunications from INRIA / University of Rennes, France. I joined Alcatel’s research and innovation center in 2000, after two years of post-doctoral study at France Telecom (Orange), where I have focused on practical aspects of my thesis applied to distributed fault diagnosis in telecommunication networks.
During the years of research at Alcatel, Alcatel-Lucent, and Nokia Bell Labs, I have been developing models and algorithms for problems of practical relevance such as autonomous network management, privacy-preserving recommender systems, causal modeling for telco and cloud networks, root cause analysis and predictive maintenance in complex distributed systems.
Currently, I am heading the Machine Learning & Systems group in AI Research Lab. The research group is composed of computer scientists, data engineers, and students that develop AI-based systems and algorithms while gaping the bridge between the promise of limitless capabilities of AI and the constraints imposed by real systems.My current research interests include deriving AI-based solutions with verifiable properties that ensure trust and provide causal interpretation of their outcomes and actions.
“…Causal questions can never be answered from data alone. They require us to formulate a model of the process that generates the data, or at least some aspects of that process.”
Judea Pearl, The Book of Why
Today’s AI approaches train the machine to associate patterns with specific meaning. This approach fails when the possibilities to learn are too many, and there is not enough data evidence to generalize over the unobserved possibilities. I believe the true intelligence of machines must be built based on the ability to reason about why we are observing a specific phenomenon and to ask counterfactual questions.
Ph.D. Signal Processing and Telecommunications from the University of Rennes and INRIA/IRISA lab, Rennes, France
M.Sc. in Industrial Engineering from the American University of Armenia, Armenia
M.Sc. and B.Sc. in Control System Engineering from Yerevan Institute of Polytechnic, Armenia
Selected articles and publications
(full list here)
Towards Building a Digital Twin of Complex System Using Causal Modelling, L. Jakovljevic, D. Kostadinov, Armen Aghasaryan, T. Palpanas. Complex Networks & Their Applications X. COMPLEX NETWORKS 2021. Springer. 2022
Reinforcement learning for antennas’ electric tilts optimization in self organizing networks, A. Massaro, D. Wellington, Armen Aghasaryan and R. Seidl. IEEE 32nd Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), 2021
Machine Learning Approaches to Early Fault Detection and Identification in NFV Architectures, A. Elmajed, Armen Aghasaryan and E. Fabre. 6th IEEE Conference on Network Softwarization (NetSoft), 2020
Resource management of cloud-enabled systems using model-free reinforcement learning, Y. Jin, M. Bouzid, D. Kostadinov, and Armen Aghasaryan. Annals of. Telecommunications 74, 2019
Common Structures in Resource Management as Driver for Reinforcement Learning: A Survey and Research Tracks, Y Jin, D. Kostadinov, M. Bouzid, Armen Aghasaryan. Machine Learning for Networking (MLN 2018). Lecture Notes in Computer Science, vol 11407. Springer, 2019
Cooperative AI-based e2e Network Slice Scaling, M.Bouzid,D.Hung Luong,D.Kostadinov; Y.Jin,L.Maggi,A.Outtagarts,Armen Aghasaryan.IEEE INFOCOM (INFOCOM WKSHPS), 2019
Stimulus-based sandbox for learning resource dependencies in virtualized distributed applications, Armen Aghasaryan, M. Bouzid and D. Kostadinov. 20th Conference on Innovations in Clouds, Internet and Networks (ICIN), 2017
Group Representation and Profiling, Armen Aghasaryan, M. Bouzid, D. Kostadinov, J. Picault. Encyclopedia of Social Network Analysis and Mining. Springer, New York, 2014.
Exploring the impact of LSH parameters in privacy-preserving personalization, Armen Aghasaryan, M. Bouzid, D. Kostadinov and A. Nandi. Bell Labs Technical Journal, vol. 18, no. 4, 2014
On the use of decentralization to enable privacy in web-scale recommendation services, A. Nandi, Armen Aghasaryan, and I. Chhabra. In Proceedings of the 12th ACM workshop on Privacy in the Electronic Society (WPES '13), 2013
On the Use of LSH for Privacy Preserving Personalization, Armen Aghasaryan, M. Bouzid, D. Kostadinov, M. Kothari and A. Nandi. 12th IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom), 2013.
Reputation Diffusion Simulation for Avoiding Privacy Violation, D. Pergament, Armen Aghasaryan and J. Ganascia. IEEE 11th International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom), 2012
P3: A Privacy Preserving Personalization Middleware for recommendation-based services., A Nandi, Armen Aghasaryan, M. Bouzid. Proc. of 4th Hot Topics in Privacy Enhancing Technologies Symposium (HotPETS 2011) at PETS 2011
Intelligent Reactive Access Control for Moving User Data, Y. Wang, Armen Aghasaryan, A. Shrihari, D. Pergament, G. Kamga and S. Betge-Brezetz. IEEE 3rd Intl. Conference on Privacy, Security, Risk and Trust and 3rd Intl. Conference on Social Computing, 2011
Privacy Data Envelope: Concept and implementation, M. Ghorbel, Armen Aghasaryan, S. Betgé-Brezetz, M. Dupont, G. Kamga and S. Piekarec. 9th Annual International Conference on Privacy, Security and Trust, 2011
Evaluation of Group Profiling Strategies, C. Senot, D. Kostadinov, M. Bouzid, J. Picault, and Armen Aghasaryan. Proceedings of the 22nd International Joint Conference on Artificial Intelligence (IJCAI), Barcelona, Spain, 2011
Multi-source user profiling and keyword inference for personalized application enablement, Armen Aghasaryan, S. Betgé-Brezetz, M. Kodialam, S. Mukherjee, C. Senot, Y. Toms, L. Wang. 13th International Conference on Intelligence in Next Generation Networks, 2009.
Ontology-Based Profiling and Recommendations for Mobile TV, Y. Naudet, Armen Aghasaryan, S. Mignon, Y. Toms, C. Senot. Semantics in Adaptive and Personalized Services. Studies in Computational Intelligence, vol 279. Springer, 2010
UML Specification of a Generic Model for Fault Diagnosis of Telecommunication Networks, Armen Aghasaryan, A., C. Jard, J. Thomas. Telecommunications and Networking - ICT 2004. Lecture Notes in Computer Science, vol 3124. Springer, 2004
Algorithms for Distributed Fault Management in Telecommunications Networks, E. Fabre, A. Benveniste, S. Haar, C. Jard, Armen Aghasaryan. Telecommunications and Networking - ICT 2004. Lecture Notes in Computer Science, vol 3124. Springer, 2004.
Fault Detection and Diagnosis in Distributed Systems: An Approach by Partially Stochastic Petri Nets, Armen Aghasaryan, E. Fabre, A. Benveniste, R. Boubour and C. Jard. Journal Discrete Event Dynamic Systems 8, 203–231, 1998
A Petri net approach to fault detection and diagnosis in distributed systems. II. Extending Viterbi algorithm and HMM techniques to Petri nets, Armen Aghasaryan, E. Fabre, A. Benveniste, R. Boubour and C. Jard. Proceedings of the 36th IEEE Conference on Decision and Control (CDC), 1997.
US-2014223575-A1 Privacy protection in recommendation services
US-2008281677-A1 System and associated method for selecting advertisements
FR-2977345-A1 DIGITAL CONTENT RECOMMENDATION SYSTEM
EP-2169854-A1 Procédé d'apprentissage de profil utilisateur et moteur de profilage associé pour plateformes combinées de fourniture de services
EP-2043010-A1 Apparatus for automatic indexing of content
EP-2270741-A1 Stratégie d'exposition personnalisée
WO-2011061457-A2Method and system for assisting with colour selection
EP-3435233-A1 A method for identifying causality objects
EP-2249592-A1 Procédé et dispositif d'obtention d'informations statistiques relatives Ã des usagers fréquentant des lieux situés dans la zone de couverture radio d'au moins un réseau de communication sans fil
FR-2963131-B1 Procédé de protection des données privées sur un réseau
WO-2013064316-A1 Système d'optimisation de circulation routière
US-2019317814-A1 Resource management of resource-controlled system
FR-2980606-A1 Method for jamming profile of user, involves transmitting additional request to service such that profile is scrambled, and validating scrambled profile in addition to additional request before transmission of service
EP-2071505-A1 Gestionnaire de communautés
FR-2834845-A1 System for creation, deployment and or monitoring of complex services implemented over a data network, e.g. video on demand, whereby services are modeled as a series of levels so that linked entities and components can be seen