Font Size: 
Fall 2021
| Dr. Gogoglou, Antonia (Facebook Inc.) |
From Thessaloniki to Silicon Valley: A Journey in Machine Learning
In this talk, I will share my journey from undergraduate studies in Electrical and Computer Engineering to Computer Science
research and finally to the tech industry. I will share my experiences in Machine Learning both in an academic and industry
setting and talk about my personal views on the challenges the field and its practitioners face.
[Tuesday] December 07, 2021 [15:30-16:30]. Talk's flyer |
| prof. Feo-Flushing, Eduardo (CMU@Doha) |
Cooperation and teamwork in the advent of artificial intelligence: Is there an elephant in the room?
The promise of artificial intelligence and robotics is to conceive technological artifacts that replicate, and even surpass, all human
capabilities. This includes distinct capabilities like locomotion, perception, speech, and learning, but must also include social
capabilities like cooperation and teamwork. Despite the advances in artificial intelligence, it is surprising that the cooperation and
teamwork capabilities of current robots fall far short of even very simple animals. So what are we missing? I will portray my views on
the importance of understanding and replicating the cognitive abilities required to cooperate in a society that includes many different
intelligent autonomous agents. How can a group of intelligent agents working together be able to solve problems that are beyond their
individual capabilities? This talk will identify and explore some of the challenges and share experiences in enabling cooperation in
heterogeneous human-robot-animal teams.
[Tuesday] November 30, 2021 [12:00-13:00]. Talk's flyer |
| Papadis, Nikolaos (PhD stud.) (YINS, Yale) |
Network optimization in blockchain and payment channel networks
Blockchain usually receives public attention whenever there is a Bitcoin price surge or due to Bitcoin's environmental impact.
As a technology, however, blockchain is much deeper: its essence and breakthrough are that it enables multiple parties to reach
consensus (e.g. on how much money one owns) without them needing to trust each other. This overview presentation aims to highlight
certain networking and network economics aspects of blockchains: how does a blockchain network normally operate and what can go wrong?
What is the role of delay on the scalability and the security of the network? What can we do to improve network throughput
(transactions per second)? We will deal with these questions and the relevant tradeoffs on the fundamental "Layer 1" network,
and then introduce the payment channel network (PCN) "Layer 2" scalability solution and optimization problems arising therein:
optimal transaction routing and scheduling, channel rebalancing, and pricing.
[Wednesday] November 24, 2021 [13:00-14:00]. Talk's flyer |
| prof. Mouratidis, Kyriakos (SMU, SG) | The talk will start with a brief introduction to the general (and widely used) top-k query, followed by observations that reveal its geometric nature. Based on the geometric aspects of the problem, we will then describe novel, auxiliary operators that complement its applications. These operators present ample computational challenges, especially for large datasets. The talk will be based on recent studies by Prof. Mouratidis, published in the ACM SIGMOD and VLDB conferences. A complete CV of the speaker, as well as full versions of the papers included in the talk, can be found at http://www.mysmu.edu/faculty/kyriakos/. |
| prof. Konstantinou, Charalambos (KAUST, SA) |
Towards Secure and Resilient Cyber-Physical Energy Systems
This talk will give an overview of the research of the Secure Next Generation Resilient Systems (SENTRY) lab (sentry.kaust.edu.sa) at
KAUST. The transformation of critical grid infrastructures into cyber-physical energy systems contributes towards modernization allowing
for better planning, more flexible control, system-wide optimization, etc. The security, however, of such systems presents significant
challenges in controlling and maintaining secure access to critical system resources and services. Cyber discovery, vulnerability
assessment, rapid risk mitigation, and resilient control of modern large-scale cyber-physical systems should consider the interdependence
between all system layers. The talk will present different cases of attack strategies simulated under nominal and abnormal operating
conditions to uncover their system-wide impacts in power systems, as well as illustrate the impact of such attacks and the feasibility
of detection methods in simulation models in order to enhance system resilience.
[Monday] October 25, 2021 [15:00-16:00]. Talk's flyer |
Spring 2021
| Dr. Kavasidis, Isaak (Univ. Catania, IT) | COVID-19 infection caused by SARS-CoV-2 pathogen has been a catastrophic pandemic outbreak all over the world, with exponential increasing of confirmed cases and, unfortunately, deaths. In this talk, an AI-powered pipeline is presented, based on the deep-learning paradigm, for automated COVID-19 detection and lesion categorization from CT scans. The system employes a segmentation module aiming at automatically identifying lung parenchyma and lobes. Next, the segmentation network is combined with classification networks for COVID-19 identification and lesion categorization. The model's classification results are compared against those obtained by three expert radiologists on a dataset of 166 CT scans. Results showed a sensitivity of 90.3% and a specificity of 93.5% for COVID-19 detection, at least on par with those yielded by the expert radiologists, and an average lesion categorization accuracy of about 84%. |
| Mr. Katsikaros, Vangelis (Adzuna, UK) |
