TY - JOUR AU - Kovásznai, Gergely AU - Kiss, D.H. AU - Mlinkó, P. TI - Formal verification for quantized neural networks JF - ANNALES MATHEMATICAE ET INFORMATICAE J2 - ANN MATH INFORM VL - 57 PY - 2023 SP - 36 EP - 48 PG - 13 SN - 1787-5021 DO - 10.33039/ami.2023.04.003 UR - https://m2.mtmt.hu/api/publication/34623343 ID - 34623343 N1 - Department of Computational Science, Eszterházy Károly Catholic University, Hungary Pázmány Péter Catholic University, Hungary Export Date: 18 February 2024 AB - Despite of deep neural networks are being successfully used in many fields of computing, it is still challenging to verify their trustiness. Previously it has been shown that binarized neural networks can be verified by being encoded into Boolean constraints. In this paper, we generalize this encoding to quantized neural networks (QNNs). We demonstrate how to implement QNNs in Python, using the Tensorflow and Keras libraries. Also, we demonstrate how to implement a Boolean encoding of QNNs, as part of our tool that is able to run a variety of solvers to verify QNNs. © 2023, Eszterhazy Karoly College. All rights reserved. LA - English DB - MTMT ER - TY - JOUR AU - Kovásznai, Gergely AU - Al-Shamarti, Mohammed TI - Integer Programming Based Optimization of Power Consumption for Data Center Networks JF - ACTA CYBERNETICA J2 - ACTA CYBERN-SZEGED PY - 2023 PG - 17 SN - 0324-721X DO - 10.14232/actacyb.299115 UR - https://m2.mtmt.hu/api/publication/34534646 ID - 34534646 AB - With the quickly developing data centers in smart cities, reducing energy consumption and improving network performance, as well as economic benefits, are essential research topics. In particular, Data Center Networks do not always run at full capacity, which leads to significant energy consumption. This paper experiments with a range of optimization tools to find the optimal solutions for the Integer Linear Programming (ILP) model of network power consumption. The study reports on experiments under three communication patterns (near, long, and random), measuring runtime and memory consumption in order to evaluate the performance of different ILP solvers.While the results show that, for near traffic pattern, most of the tools rapidly converge to the optimal solution, CP-SAT provides the most stable performance and outperforms the other solvers for the long traffic pattern. On the other hand, for random traffic pattern, Gurobi can be considered to be the best choice, since it is able to solve all the benchmark instances under the time limit and finds solutions faster by 1 or 2 orders of magnitude than the other solvers do. LA - English DB - MTMT ER - TY - BOOK ED - Gazdag, Zsolt ED - Iván, Szabolcs ED - Kovásznai, Gergely TI - Proceedings of the 16th International Conference on Automata and Formal Languages PB - Electronic Proceedings in Theoretical Computer Science (EPTCS) PY - 2023 SP - 284 DO - 10.4204/EPTCS.386 UR - https://m2.mtmt.hu/api/publication/34439161 ID - 34439161 LA - English DB - MTMT ER - TY - JOUR AU - Kovásznai, Gergely AU - Varga, Imre TI - Special Issue on Applied Informatics JF - INFOCOMMUNICATIONS JOURNAL J2 - INFOCOMM J VL - 15 PY - 2023 IS - Special Issue SP - 1 EP - 1 PG - 1 SN - 2061-2079 UR - https://m2.mtmt.hu/api/publication/34131002 ID - 34131002 LA - English DB - MTMT ER - TY - JOUR AU - Al-Shamarti, Mohammed AU - Kovásznai, Gergely AU - Malik, Ali AU - de Fréin, Ruairí TI - Survey of Routing Techniques-Based Optimization of Energy Consumption in SD-DCN JF - INFOCOMMUNICATIONS JOURNAL J2 - INFOCOMM J VL - 15 PY - 2023 IS - Special Issue SP - 35 EP - 42 PG - 8 SN - 2061-2079 DO - 10.36244/ICJ.2023.5.6 UR - https://m2.mtmt.hu/api/publication/34130917 ID - 34130917 N1 - Department of Information Technology, University of Debrecen, Hungary Department of Computational Science, Eszterházy Károly Catholic University, Eger, Hungary School of Electrical and Electronic Engineering, Technological University, Dublin, Ireland Export Date: 18 February 2024 Funding details: Science Foundation Ireland, SFI, 13/ RC/2077 P2, 15/SIRG/3459 Funding text 1: This research was supported by the department of the Information Technol-This research was supported by the department of the Information Technology, University of Debrecen. This paper has emanated from research supported ogy, University of Debrecen. This paper has emanated from research supported in part by a Grant from Science Foundation Ireland under Grant numbers in13/RC/2077part by aP2Grantandfrom15/SIRG/3459.Science Foundation Ireland under Grant numbers in part by a Grant from Science Foundation Ireland under Grant numbers 13/ Department of Information Technology, University of Debrecen, Hungary. Department of Information Technology, University of Debrecen, Hungary. Department of Information Technology, University of Debrecen, Hungary. Department of Computational Science, Eszterházy Károly Catholic Univer-Department of Computational Science, Eszterházy Károly Catholic Univer-Department of Computational Science, Eszterházy Károly Catholic Univer- 3sity, Hungary. Email: kovasznai.gergely@uni-eszterhazy.hu; School of Electrical and Electronic Engineering, Technological University School of Electrical and Electronic Engineering, Technological University School of Electrical and Electronic Engineering, Technological University Dublin, Ireland. Email (ali.malik, ruairi.defrein)@tudublin.ie. Dublin, Ireland. E-mail (ali.malik, ruairi.defrein)@tudublin.ie. Funding text 2: This research was supported by the department of the Information Technology, University of Debrecen. This paper has emanated from research supported in part by a Grant from Science Foundation Ireland under Grant numbers 13/ RC/2077 P2 and 15/SIRG/3459. The authors would like to thank the anonymous peer reviewers for their valuable insights and constructive comments. Also, we thank the Department of Information Technology at the University of Debrecen, and the Science Foundation Ireland (SFI) for supporting this work. Funding text 3: ACKNOWLEDGMENTS The authors would like to thank the anonymous peer reviewers for their valuable insights and constructive comments. Also, we thank the Department of Information Technology at the University of Debrecen, and the Science Foundation Ireland (SFI) for supporting this work. AB - The increasing power consumption of Data Center Networks (DCN) is becoming a major concern for network operators. The object of this paper is to provide a survey of state-of-the-art methods for reducing energy consumption via (1) enhanced scheduling and (2) enhanced aggregation of traffic flows using Software-Defined Networks (SDN), focusing on the advantages and disadvantages of these approaches. We tackle a gap in the literature for a review of SDN-based energy saving techniques and discuss the limitations of multi-controller solutions in terms of constraints on their performance. The main finding of this survey paper is that the two classes of SDNbased methods, scheduling and flow aggregation, significantly reduce energy consumption in DCNs. We also suggest that Machine Learning has the potential to further improve these classes of solutions and argue that hybrid ML-based solutions are the next frontier for the field. The perspective gained as a consequence of this analysis is that advanced ML-based solutions and multi-controller-based solutions may address the limitations of the state-of-the-art, and should be further explored for energy optimization in DCNs. LA - English DB - MTMT ER - TY - CHAP AU - Al-Shamarti, Mohammed AU - Kovásznai, Gergely AU - Abboosh, M. AU - Malik, A. AU - Frein, R.D. ED - Fazekas, István TI - ML-Based Online Traffic Classification for SDNs T2 - 2022 IEEE 2nd Conference on Information Technology and Data Science (CITDS) PB - Institute of Electrical and Electronics Engineers (IEEE) CY - Piscataway (NJ) SN - 9781665496520 PY - 2022 SP - 217 EP - 222 PG - 6 DO - 10.1109/CITDS54976.2022.9914138 UR - https://m2.mtmt.hu/api/publication/33293087 ID - 33293087 N1 - University of Debrecen, Department of Information Technology, Hungary Eszterházy Károly, Catholic University, Department of Computational Science, Hungary University of Debrecen, Department of Data Science and Visualization, Hungary AB - Traffic classification is a crucial aspect for Software-Defined Networking functionalities. This paper is a part of an on-going project aiming at optimizing power consumption in the environment of software-defined datacenter networks. We have developed a novel routing strategy that can blindly balance between the power consumption and the quality of service for the incoming traffic flows. In this paper, we demonstrate how to classify the network traffic flows so that the quality of service of each flow-class can be guaranteed efficiently. This is achieved by creating a dataset that encompasses different types of network traffic such as video, VoIP, game and ICMP. The performance of a number of Machine Learning techniques is compared and the results are reported. As part of future work, we will incorporate classification into the power consumption model towards achieving further advances in this research area. © 2022 IEEE. LA - English DB - MTMT ER - TY - JOUR AU - Al-Shamarti, Mohammed AU - Kovásznai, Gergely AU - Rácz, Anett AU - Malik, Ali AU - de Fréin, Ruairí TI - An Adaptive Routing Framework for Efficient Power Consumption in Software-Defined Datacenter Networks JF - ELECTRONICS (SWITZ) VL - 10 PY - 2021 IS - 23 SP - 3027 SN - 2079-9292 DO - 10.3390/electronics10233027 UR - https://m2.mtmt.hu/api/publication/32545784 ID - 32545784 N1 - Department of Information Technology, University of Debrecen, Debrecen, 4032, Hungary Department of Computational Science, Eszterhazy Karoly Catholic University, Eger, 3300, Hungary Department of Applied Mathematics & Probability Theory, University of Debrecen, Debrecen, 4032, Hungary School of Electrical and Electronic Engineering, Technological University Dublin, Dublin, D07 EWV4, Ireland Cited By :2 Export Date: 6 December 2022 Correspondence Address: de Fréin, R.; School of Electrical and Electronic Engineering, Ireland; email: ruairi.defrein@tudublin.ie Funding details: Science Foundation Ireland, SFI, 15/SIRG/3459 Funding text 1: Funding: This publication has emanated from research conducted with the financial support of Science Foundation Ireland (SFI) under the Grant Number 15/SIRG/3459. LA - English DB - MTMT ER - TY - JOUR AU - Kovásznai, Gergely AU - Gajdár, Krisztián AU - Narodytska, Nina TI - Portfolio solver for verifying Binarized Neural Networks JF - ANNALES MATHEMATICAE ET INFORMATICAE J2 - ANN MATH INFORM VL - 53 PY - 2021 SP - 183 EP - 200 PG - 18 SN - 1787-5021 DO - 10.33039/ami.2021.03.007 UR - https://m2.mtmt.hu/api/publication/32169794 ID - 32169794 N1 - Export Date: 6 December 2022 Correspondence Address: Kovásznai, G.; Eszterházy Károly UniversityHungary; email: kovasznai.gergely@uni-eszterhazy.hu LA - English DB - MTMT ER - TY - JOUR AU - Kovásznai, Gergely AU - Fazekas, I. AU - Tómács, Tibor TI - Preface JF - CEUR WORKSHOP PROCEEDINGS J2 - CEUR WORKSHOP PROC VL - 2650 PY - 2020 SN - 1613-0073 UR - https://m2.mtmt.hu/api/publication/32842987 ID - 32842987 LA - English DB - MTMT ER - TY - GEN AU - Kovásznai, Gergely TI - Szenzorhálózatok optimalizálása formális módszerekkel PY - 2020 UR - https://m2.mtmt.hu/api/publication/32169856 ID - 32169856 LA - English DB - MTMT ER -