Unige Instructors | ||
Computational models of visual perception | Fabio Solari | 13-17 Feb 2023 |
Theory and Practice of Virtual Reality Systems | Manuela Chessa | 27 Feb -3 Mar 2023 |
Introduction to Type Theory: from foundations to practice | Francesco Dagnino | 6-10 Mar 2023 |
Introduction to formal verification: an appetiser | Angelo Ferrando, Giorgio Delzanno | 13-17 Mar 2023 |
Strategic Choices: Games and Team Optimization | Marcello Sanguineti Lucia Pusillo | 27-31 Mar 2023 |
An Introduction to Prolog | Viviana Mascardi | 3-5 Apr 2023 |
Ricca - Leotta school May 2023 | Filippo Ricca, Maurizio Leotta | May 2023 |
Mobile Security | Alessio Merlo | 22-26 May 2023 |
Introduction to High Performance Computing | Daniele D'Agostino | 29 May - 2 Jun |
Theory and Practice of Runtime Monitoring | Davide Ancona | 5-9 Jun 2023 |
CVCC + DL | Francesca Odone Nicoletta Noceti | 5-9 Jun 2023 |
Machine Learning School | Lorenzo Rosasco | 19-23 Jun 2023 |
Effective habits and skills for successful young scientists | Fabio Roli | 26-30 Jun 2023 |
Adversarial Machine Learning | Fabio Roli, Luca Demetrio | 3-5 Jul 2023 |
Trustworthy Artificial Intelligence | Luca Oneto | 10-14 Jul 2023 |
Distributed Optimization and multi decision making | Giulio Ferro, Michela Robba | 24-27 luglio 2023 |
External Instructors (CNR/IMT Lucca) | ||
An introduction to optimization over time and its application to online machine learning and reinforcement learning | Giorgio Gnecco | 20-24 Feb 2023 |
An Introduction to Model Predictive Control and Rolling Horizon Optimization | Mauro Gaggero | 20-24 Mar 2023 |
Network monitoring and inspection | Matteo Repetto | 27-31 Mar 2023 |
Accelerated Parallel Systems: the GPU and FPGA cases | Antonella Galizia | 15-19 May 2023 |
Information Hiding | Luca Caviglione | 17-21 Jul 2023 |
Detailed information
Computational models of visual perception
Duration: 20 hours (+ final project)
Instructor:
Fabio Solari – DIBRIS, University of Genoa – fabio.solari@unige.it
When: 13-17 February 2023
Where: via Dodecaneso 35
Abstract
This course introduces paradigms and methods that allow students to develop computational models of visual perception, which are based on hierarchical networks of interacting neural units, mimicking biological processing stages.
Program
References
An introduction to optimization over time and its application to online machine learning and reinforcement learning
Duration: 20 hours
Instructor:
Giorgio Gnecco – IMT Lucca – giorgio.gnecco@imtlucca.it
When: 20 -24 February 2023
Where: In one of the DIBRIS buildings either in Via Dodecaneso or in Via Opera Pia. Students’ attendance on Teams is also possible.
Abstract
The course first provides an introduction to classical methods of dynamic optimization, such as the Bellman Optimality Principle and the Pontryagin Principle. Then, it introduces more recent topics, such as Approximate Dynamic Programming and Neural Networks for the approximate solution of dynamic optimization problems. The proofs are outlined; the details are presented only when they provide useful insights. Both discrete-time and continuous-time optimization are considered. Several applications and case-studies are described and discussed. MATLAB code will be presented for some examples.
Program
Discrete-time N-stage optimization: Dynamic Programming, LQ and LGQ problems, The Riccati Equations, The Kalman Filter, Approximate Dynamic Programming, Neural Networks for approximate solutions to discrete-time N-stage optimization problems. Continuous-time optimization: the Hamilton-Jacobi-Bellman Equation and the Pontryagin Principle. Connection between continuous-time optimization and differential games. Case-studies and examples.
References
- D. P. Bertsekas: “Dynamic Programming and Optimal Control”, vol.I, Athena Scientific, fourth edition, 2017.
- D. P. Bertsekas: “Dynamic Programming and Optimal Control”, vol. II, Athena Scientific, fourth edition, 2012.
- Lecture notes provided by the teacher.
Theory and Practice of Virtual Reality Systems
Duration: 20 hours
Instructor(s): Manuela Chessa - University of Genoa (DIBRIS) - manuela.chessa@unige.itThis email address is being protected from spambots. You need JavaScript enabled to view it.">
When: 27 February - 3 March 2023
Where: via Dodecaneso 35, Valletta Puggia, DIBRIS
Abstract
The course provides a general introduction to the theory and the development of Virtual Reality Systems. The course will start from some basic aspects of Virtual Reality towards the recent achievements in Mixed Reality. The course will cover the following topics.
References
Introduction to Type Theory: from foundations to practice
Duration: ~24 hours (about 20 hours) 20 hours for lectures + 4 hours for exercises/laboratory
Instructor(s):
Francesco Dagnino – DIBRIS, UniGe – francesco.dagnino@dibris.unige.it
Jacopo Emmenegger – DIMA, UniGe – emmenegger@dima.unige.it
When: 6-10 March 2023
Where: DIBRIS @ VP, Genova
Abstract
Proof assistants are tools designed to write formal proofs and automatically check their correctness. They are increasingly used in many different domains, from software verification to mathematics. They allow, for instance, to write code correct-by-construction or to formalise complicated mathematical arguments.. Most popular proof assistants, such as Agda, Coq or Lean, implement a constructive logic based on a (dependent) type theory. This means that they are strongly typed functional programming languages where types and programs are seen as logical formulas and proofs, respectively, and then correctness is just ensured by type-checking a program.
In the course, we will study fundamental notions and results on type theories, explaining their connection with logic, and we will experiment formal reasoning in a type theory, using Agda as a concrete system.
Program
The program can be adapted depending on the audience.
- The untyped lambda-calculus as a computational model: terms, reduction, confluence, normalisation
- Constructive reasoning, Intuitionistic Propositional Logic and the BHK interpretation
- Simply typed lambda-calculus, proposition-as-types, proofs-as-terms
- Strong normalisation and consistency
- Introduction to dependent types, quantifiers via dependent sums and products
- Dependent types in Agda, universes
- Inductive types in Agda
- Equality via Identity types
References
[1] J.Y. Girard, Y. Lafont, P. Taylor. Proofs and Types. Cambridge University Press, 1989.
21] M.H.B. Sorensen, P. Urzyczyn. Lectures on the Curry-Howard Isomorphism. Elsevier, 2006.
[3] B. Nordstrom, K. Petersson, J.M. Smith. Programming in Martin-löf’s type theory : an introduction. Clarendon Press, 1990.
[4] M. Hofmann. Syntax and Semantics of Dependent Types. Cambridge University Press, 1997
[5] Agda (https://agda.readthedocs.io/en/v2.6.2/)
Introduction to formal verification: an appetiser
Duration: about 20 hours
Instructor(s): Angelo Ferrando – University of Genova - angelo.ferrando@unige.it
Giorgio Delzanno - University of Genova
When: 13-17 March 2023
Where: Via Dodecaneso 35, Valletta Puggia, DIBRIS
Abstract
The course provides a general introduction to static formal verification (such as Model Checking), and runtime verification. The course will not focus only on the theoretical foundations of the two approaches, but it will offer practical insights as well. For both methodologies, established tools will be presented and experimented (through laboratories). At the end of the course, a more general overview of recent works on runtime verification will also be reported. This will help the students to have a better understanding of the newest features and challenging applications where such formal technique has been applied.
Program
References
An Introduction to Model Predictive Control
and Rolling Horizon Optimization
Duration: 20 hours
Instructor: Mauro Gaggero, National Research Council of Italy (CNR), Genova - mauro.gaggero@cnr.it
When: 20-24 Marzo 2023
Where: University of Genoa (classroom to be announced in Via Opera Pia) or Microsoft Teams platform
Abstract
Model predictive control (MPC) and rolling-horizon optimization are optimization and control paradigms that have been widely employed in the literature owing to their ability to exploit information on the future behavior of the system at hand, their capability of dealing with constraints, and the presence of many theoretical results about their properties. From various decades, MPC has been used for process control in chemical plants, and nowadays it is employed for the optimization of many other complex setups such as, for instance, power plants, mechatronic systems, logistics operations, cloud computing applications, and so on. It is still receiving on-going interest from researchers in both the industrial and academic communities. Concerning the academic world, MPC is attractive for both researchers working in the field of Control Systems and Operations Research since it combines several aspects of both disciplines. The course will start from the basic theoretic notions of MPC and rolling-horizon optimization, together with recent developments in design and implementation. Special attention will be devoted to the computational aspects of MPC and to the existing techniques to reduce the overall required effort. An overview of receding-horizon state estimation, a topic strictly related to MPC, will be given as well. Finally, recent applications of MPC and rolling-horizon optimization will be presented, together with details of their software implementation.
Program
References
Strategic Choices: Games and Team Optimization
Duration: 20 hours
Instructors: Lucia Pusillo - University of Genoa (DIMA) - pusillo@dima.unige.it
Marcello Sanguineti - University of Genoa (DIBRIS) - marcello.sanguineti@unige.it
When: 27-31 March 2023
Where: Via Dodecaneso 35
Abstract: Game and Team Theory study strategic interactions among two or more agents, which have to take decisions in order to optimize their objectives. They have various links to disciplines such as Economics, Engineering, Computer Science, Political and Social Sciences, Biology, and Medicine. These links provide incentives for interdisciplinary research and make the role of Game and Team Theory invaluable in a variety of applications. The main goal of this course consists in providing students with the basic mathematical tools to deal with interactive problems and illustrating them via case-studies.
Program
References
Network monitoring and inspection
Duration: 20 hours
Instructor(s):
Matteo Repetto
Institute for Applied Mathematics and Information Technologies
National Research Council of Italy (CNR)
matteo.repetto@ge.imati.cnr.it
When: 27-31 March 2023
Where: Online (Teams) or on site (as requested by students)
Abstract
The Internet is the main carrier for cyber-attacks, so it is not surprising that most detection techniques build on network flow monitoring and packet inspection. There is a huge amount of information that potentially can be gathered from the network, but deep packet inspection at line rate is extremely challenging even in hardware, especially in case of high-speed links (1 Gbps and upward). This course will give a basic understanding of common tools for flow monitoring and packet inspection, with specific emphasis on how to extract custom information that is ever more needed to detect modern attacks. Besides, the eBPF framework provided by the Linux kernel will be introduced as a power mechanism to build efficient, custom and portable inspection and enforcement processes.
Program
References
An Introduction to Prolog
Duration: 12 hours
Instructor: Mascardi Viviana, viviana.mascardi@unige.it
When: April, 3, 4, 5, 2023 (4 hours per day, given in presence and online via Teams)
Where: DIBRIS, Via Dodecaneso 35
Abstract: "In the summer of 1972, Alain Colmerauer and his team in Marseille developed and implemented the first version of the logic programming language Prolog. Together with both earlier and later collaborations with Robert Kowalski and his colleagues in Edinburgh, this work laid the practical and theoretical foundations for the Prolog and logic programming of today. Prolog and its related technologies soon became key tools of symbolic programming and Artificial Intelligence."
This statement is taken from the "The Year of Prolog" web page, http://prologyear.logicprogramming.org/: the 50th anniversary of Prolog has been celebrated in 2022 with scientific and dissemination initiatives all over the world. Keeping the momentum going, we offer a compact introductory course on Prolog, with practical exercises and an overview of the existing and future Prolog applications. In particular, we highlight the potential of Prolog for implementing cognitive intelligent agents and for supporting eXplainable Artificial Intelligence (XAI) thanks to its declarative flavor.
Program:
Prolog syntax
Prolog operational semantics
Extra-logical and meta-logical predicates
Examples
Applications
Future perspectives
References:
The Art of Prolog, second edition
Advanced Programming Techniques
by Leon S. Sterling and Ehud Y. Shapiro
1994
Accelerated Parallel Systems: the GPU and FPGA cases
Duration: 20 hours (five half-days)
Instructors: Antonella Galizia - IMATI-CNR antonella.galizia@ge.imati.cnr.it, Christian Pilato - Politecnico di Milano christian.pilato@polimi.itThis email address is being protected from spambots. You need JavaScript enabled to view it.">
When: 15-19 May 2023
Where: DIBRIS-UNIGE Via Dodecaneso 35, Genova (online attendance will be possible)
Abstract
With the end of Moore law for sequential computing architectures and the advents of multi and many cores era, managing parallelism is no longer the goal of a restricted community but becomes a need for everybody who is interested in exploiting an adequate fraction of available performance provided by widespread modern computing architectures, including desktop and mobile devices. A computer is nowadays a complex system with heterogeneous computational units including multi cores CPU and many cores accelerators such as Graphic Processing Unit (GPU), Field-Programmable Gate Array (FPGA), or others. The aim of the course is to present the state-of-the-practice on computing systems equipped with accelerated technology; the main focus is on high efficiency, which is of utmost importance but can have different meanings: as for high-performance computing and data center domains, high efficiency mostly relates to performance while in the mobile and IoT space, research communities think about accelerators more from a power/energy perspective. The course considers programming of Complex Heterogeneous Parallel Systems (CHPS) and in particular accelerators as GPU and FPGA, the overall goal and challenge is the portability and performances of software to ensure effectiveness and efficiency of target applications. This edition of the course will discuss two different approaches: the GPGPU based solutions and the hardware specialization of the application on FPGA. In particular, it will be shown, with practical cases, how to design and implement applications able to exploit available computational resources through a suitable selection of programming tools, communications and domain-oriented libraries, and design and implementation strategies. At this regards the course will include a hands-on part that the student may dedicate to a general case study or to a personalized case depending on specific interests.
Program
References
Slides of the course will be provided to students
Ricca - Leotta school May 2023
Mobile Security
Duration: 20 hours (5 half-days)
Instructors(s): Alessio Merlo – University of Genova alessio.merlo@unige.it Luca Verderame – University of Genova luca.verderame@unige.itThis email address is being protected from spambots. You need JavaScript enabled to view it.">
When: 22-26 May 2023
Where: online via Teams or in presence at DIBRIS – Valletta Puggia, Via Dodecaneso 35
Abstract: The course provides an overview of the main topics related to the security of mobile devices and applications. The course offers an insight into the leading mobile operating systems (i.e., Android and iOS) and their security issues. Moreover, the course provides a discussion on emerging mobile security technologies (e.g., Host-Based Card Emulation, and Trusted Execution Environment), security threats, and possible countermeasures. The second part of the course will cover the security of Mobile Applications, with a particular focus of state-of-the-art frameworks and methodologies for the vulnerability assessment of Android applications. Finally, the course will provide specific hands-on sessions with tools and techniques for the vulnerability assessment of Android applications.
Program:
During the hands-on sessions, students will get acquainted with a number of static and dynamic analysis tools, including
References:
Introduction to High Performance Computing
Duration: 20 hours
Instructor: Daniele D’Agostino – DIBRIS Unige
When: 29 May - 2 June 2023
Where: Via Dodecaneso
Abstract For most scientists the abstract fact of the existence of an algorithm solving a problem is enough, while its efficient implementation in terms of exploitation of the available computational capabilities is mostly disregarded. But with the end of Moore law for sequential computing architectures and the advents of multi and many cores era, managing parallelism is no longer the goal of a restricted ICT community, it becomes a need for everybody who is interested in exploiting an adequate fraction of available performance provided by widespread modern computing architectures. The aim of the course is to provide a glance of the different aspects involved in efficient and effective programming of current heterogeneous computing systems equipped with manycore x86 architectures and accelerators, in particular graphics cards (GPUs). Therefore, it conveys the required knowledge to develop a thorough understanding of the interactions between software and hardware at the core, socket, node and cluster level. In particular it will be presented, with practical cases, how the design and implementation of programs can exploit available computational resources through a suitable selection of programming paradigms, compiling and profiling tools. The course includes a hands-on part that the student may dedicate to a general case study or to a personalized case depending on specific interests.
The programming language will be C/C++. The translation in Fortran is straightforward.
Program
References
Theory and Practice of Runtime Monitoring
Duration: about 20 hours
Instructor(s): Davide Ancona - University of Genoa (DIBRIS) - davide.ancona@unige.it,
Angelo Ferrando - University of Genoa (DIBRIS) - angelo.ferrando@unige.it
When: 5 - 9 June 2023
Where: via Dodecaneso 35, Valletta Puggia, DIBRIS
Abstract
The course provides a general introduction to Runtime Monitoring and Verification (RM&V), and the theoretical and practical aspects of RML (Runtime Monitoring Language), a system agnostic domain specific language for RM&V. Use cases will be considered in the context of distributed and Internet of Things systems with Node.js/Jalangi2 and Node-RED and robotic systems based on ROS.
References
Computer Vision Crash Course
Duration: 20 hours
Instructor(s):
Francesca Odone and Nicoletta Noceti
MaLGa DIBRIS, Università degli Studi di Genova
{francesca.odone, nicoletta.noceti,}@unige.it
When: 5 -9 June 2023
Where: DIBRIS-UNIGE Via Dodecaneso 35, Genova
Abstract
Visual perception, as a key element of Artificial Intelligence, allows us to build smart systems sensitive to surrounding environments, interactive robots, video-cameras with real time algorithms running on board. With similar algorithms, our smart phones can log us in by recognizing our face, read text automatically, improve the quality of the photos we shoot. At the core of these applications are computer vision models, often boosted by machine learning algorithms.
This crash course is conceived as a complement to the “Deep Learning: Hands on introduction” course (henceforth DL) although it can be taken independently.
It covers the basic principles of computer vision and visual perception in artificial agents, including theoretical classes, application examples, hand-on activities.
Within CVCC, we present elements of classical computer vision (introduction to image processing, feature detection, depth estimation, motion analysis).
At the same time, by borrowing from DL, we also present deep learning approaches to computer vision problems such as image classification, detection, and semantic segmentation.
Core CVCC Program (for those attending the CVCC course only)
Integrated DL and CVCC program
References
Slides and readings will be provided.
Some reference books:
Topics in Modern Machine Learning (ModML)
Duration
20 hours
Instructor(s)
Lorenzo Rosasco – DIBRIS – lorenzo.rosasco@unige.itThis email address is being protected from spambots. You need JavaScript enabled to view it.">
Giovanni Alberti – DIMA – giovanni.alberti@unige.it
Simone Di Marino – DMA - simone.dimarino@unige.it
When: 19-23 June 2023
Where DIBRIS, Via Dodecaneso 35
Abstract
This is an advanced machine learning course covering some of the topics of interest in modern machine learning. After a 6-hour boot camp on the first day (on statistical learning, machine learning models and optimization for machine learning), the rest of the week will be dedicated to introducing modern topics including, for example, interpolation and overparameterization, implicit regularization, optimal transport for machine learning, machine learning for inverse problems, fairness in machine learning and reinforcement learning. Each presentation, held in the morning, will be introductory and self-contained, with an associated practical session in the afternoon. The last day will be dedicated to a workshop with invited speakers.
Program The detailed program will be provided later.
References Ad hoc references will be given for each topic.
Effective habits and skills for successful young scientists
Teacher: Fabio Roli
Duration: 20 hours (5 half-days)
When: June 26-30 2023
Where: online on MS Teams
Curriculum: Cross-curricula course
Exam: written assessments with open-ended questions
Abstract:
Although tons of books on effective habits and soft skills have been published, they have not been thought for scientists, and, therefore, issues that are relevant for them are not easily available. This short course aims to collect spread ideas and place them in a coherent framework useful for young scientists and provide a small tactical guide for scientists at the first stages of their career. First, I review the main concepts of Steve Covey's personal and time management paradigm, the inspirational speeches of the late Professor Randy Pausch, and the paradigm of atomic habits of James Clear, and discuss their utility for daily activity of a young scientist. Then, I focus on a few practical skills, namely, on how to write a great paper and give a great talk. I try to convey the message that succeeding in science and technology requires skills and habits beyond the pure intelligence and intellectual abilities, and that good habits and skills of personal and time management are extremely important for young scientists.
Program:
- Basic concepts of theory of habits. Effective habits for young scientists.
- Basis concepts of personal and time management. Effective personal and time management for young scientists.
- Survival skills in the game of science. Know yourself: match your goals to your character and talents.
- How to write a great paper.
- How to give a great talk.
References:
Adversarial Machine Learning
Teachers: Fabio Roli and Luca Demetrio
Duration: 12 hours (3 half-days)
When: July 3-5 2023
Where: online on MS Teams
Curriculum: Cybersecurity and Reliable AI
Exam: 2 written assessments (one with multiple choice questions, one hands-on assessment using the SecML software library, https://secml.readthedocs.io/en/v0.15/ )
Abstract:
Today machine-learning algorithms are used for many real-world applications, including image recognition, spam filtering, malware detection, biometric recognition. In these applications, the learning algorithm can have to face intelligent and adaptive attackers who can carefully manipulate data to purposely subvert the learning process. As machine learning algorithms have not been originally designed under such premises, they have been shown to be vulnerable to well-crafted attacks, including test-time evasion and training-time poisoning attacks (also known as adversarial examples). In particular, the security of cloud-based machine-learning services has been questioned through the careful construction of adversarial queries that can reveal confidential information on the machine-learning service and its users. This course aims to introduce the fundamentals of the security of machine learning, the related field of adversarial machine learning, and some techniques to assess the vulnerability of machine-learning algorithms and to protect them from adversarial attacks. We report application examples including object recognition in images, biometric identity recognition, spam and malware detection, with hands-on on attacks against machine learning and defences of machine-learning algorithms using the SecML software library, https://secml.readthedocs.io/en/v0.15/.
Program:
- Introduction to adversarial machine learning: introduction by practical examples from computer vision, biometrics, spam, malware detection.
- Design of learning-based pattern classifiers in adversarial environments. Modelling adversarial tasks. The two-player model (the attacker and the classifier). Levels of reciprocal knowledge of the two players (perfect knowledge, limited knowledge, knowledge by queries and feedback). The concepts of security by design and security by obscurity
- System design: vulnerability assessment and defense strategies. Attack models against machine learning. Vulnerability assessment by performance evaluation. Taxonomy of possible defense strategies.
- Hands-on classes on attacks and defences of machine-learning algorithms using the SecML open-source Python library for the security evaluation of machine learning algorithms (https://secml.readthedocs.io/en/v0.15/ ).
- Summary and outlook. Current state of this research field and future perspectives
References:
Trustworthy Artificial Intelligence
Duration: 20 hours
Instructor(s): Luca Oneto – UNIGE – luca.oneto@unige.it
When: from Monday 10th of July 2023 to Friday 14th of July 2023 from 8:00 to 12:00 CEST
Where: TBD
Abstract: It has been argued that Artificial Intelligence (AI) is experiencing a fast process of commodification. This characterization is of interest for big IT companies, but it correctly reflects the current industrialization of AI. This phenomenon means that AI systems and products are reaching the society at large and, therefore, that societal issues related to the use of AI and Machine Learning (ML) cannot be ignored any longer. Designing ML models from this human-centered perspective means incorporating human-relevant requirements such as reliability, fairness, privacy, and interpretability, but also considering broad societal issues such as ethics and legislation. These are essential aspects to foster the acceptance of ML-based technologies, as well as to be able to comply with an evolving legislation concerning the impact of digital technologies on ethically and privacy sensitive matters.
Program
References
Information Hiding
Duration: 20 hours (5 half-days)
Instructor:
Luca Caviglione
Institute for Applied Mathematics and Information Technologies
National Research Council of Italy (CNR)
luca.caviglione@ge.imati.cnr.it
When: 17-21 Jul 2023
Where: preferred venue is via Skype or Microsoft Teams to reach a wide audience. Otherwise, Via Dodecaneso (according to post-pandemic rules).
Abstract:
Information hiding techniques are increasingly used in investigative journalism to protect the identity of sources or by malware to hide its existence and communication attempts. Therefore, understanding how information hiding can be used to empower privacy of users or endow malicious software with the ability of staying "under the radar" are essential to fully assess the modern cybersecurity panorama.
In this perspective, the course introduces the use of information hiding in modern threats and privacy-enhancing architectures with emphasis on two different research areas, specifically: i) techniques for creating network covert channels for communicating with a remote command & control facility, exfiltrate sensitive information and or enforce privacy ii) how to create and detect a covert channel implementing an abusive local path between two colluding applications to bypass the security framework of mobile devices.
To give a comprehensive overview on information hiding and steganography, the course will also cover the use of information hiding and steganographic techniques for watermarking purposes. For instance, it will showcase the main mechanisms for watermarking images, sounds and network flows for management, retrieval, metadating, authentication and copyright enforcement. The course will also discuss possible countermeasures or mitigation methodologies for facing the risks of the increasing amount of steganographic threats observed in the wild.
Program:
Module 1: Course introduction and a general view on information hiding.
Module 2: Information hiding as a cybersecurity threat: malware and colluding applications.
Module 3: Network covert channels (including air-gapped covert channels).
Module 4: Information hiding for watermarking, privacy enhancing, and metadating.
Module 5: Countermeasures (e.g., detecting obfuscated malware or removing ambiguities in protocols).
References:
[1] W. Mazurczyk, L. Caviglione, “Steganography in Modern Smartphones and Mitigation Techniques”, IEEE Communications Surveys & Tutorials, IEEE, Vol. 17, No.1, First Quarter 2015, pp. 334 - 357.
[2] W. Mazurczyk, L. Caviglione, Information Hiding as a Challenge for Malware Detection, IEEE Security & Privacy, Vol. 13, No. 2, pp. 89-93, Mar.-Apr. 2015.
[3] L. Caviglione, M. Podolski, W. Mazurczyk, M. Ianigro, “Covert Channels in Personal Cloud Storage Services: the case of Dropbox”, IEEE Transactions on Industrial Informatics, IEEE, Vol. 13, No. 4, pp. 1921 - 1931, August 2017.
[4] L. Caviglione, M. Gaggero, J.-F. Lalande, W. Mazurczyk, M. Urbanski, “Seeing the Unseen: Revealing Mobile Malware Hidden Communications via Energy Consumption and Artificial Intelligence”, IEEE Transactions on Information Forensics & Security, IEEE, Vol. 11, No. 4, pp. 799 – 810, April 2016.
[5] W. Mazurczyk, L. Caviglione, “Cyber Reconnaissance Techniques”, Communications of the ACM, Vol. 64, No. 3, pp. 86-95, March 2021.
[6] L. Caviglione, W. Mazurczyk, “Never Mind the Malware, Here’s the Stegomalware”, IEEE Security & Privacy, Vol. 20, No. 5, pp. 101-106, Sept.-Oct. 2022.
[7] Steg-in-the-wild (https://github.com/lucacav/steg-in-the-wild): a curated list of attacks observed in the wild taking advantage of steganographic or information-hiding-capable techniques.
Distributed optimization and multi decision making
Duration: 20 hours
Instructors: Giulio Ferro– University of Genova – giulio.ferro@unige.it; Michela Robba– University of Genova – michela.robba@unige.it
The course will include the participation of personnel from MIT (Massachusetts Institute of Technology) for research activities related to the course.
When: 24-27 July 2022
Where: Teams and in presence
Abstract
In recent years, there is a growing attention on distributed optimization due to recent interest in large scale optimization problems application, like machine learning smart grids and general networked systems, involving multiple decision makers at the same time. In this context, distributed optimization techniques are a very promising solution to afford this type of decision problems, since allow the fast solution of large-scale optimization problems by keeping private the information exchange and allowing plug and play capabilities. The course will present the basics of distributed optimization and multi decision making starting from the basic concepts of constrained optimization, up to developing and implementing state of the art distributed optimization algorithms (such as dual decomposition and ADMM). From an application point of view, the course will be mainly focused on examples related to energy management (smart grids, energy communities, electric vehicles). In some cases, the examples will regard coupled transportation and energy networks. Finally, a seminar will be held on recent developments in distributed optimization (with application to smart grids) by experts from the Massachusetts Institute of Technology (MIT), with the instructors have been working for years.
Program
- Introduction to constrained optimization and applications in the energy and transportation sector (electric vehicles)
- Multi decision making with multilevel programming
- Distributed optimization algorithms: dual decomposition and alternating direction method of multipliers (ADMM).
- Seminar on recent developments on distributed optimization by MIT experts.
- Application to real case studies: smart grids, energy communities, electric vehicles.
References
- Romvary J, Ferro G, Haider R, Annaswamy A. A Proximal Atomic Coordination Algorithm for Distributed Optimization. IEEE Trans Automat Contr. 2021:1-1. doi:10.1109/tac.2021.3053907
- Ferro G, Robba M, D’Achiardi D, Haider R, Annaswamy A. A distributed approach to the Optimal Power Flow problem for unbalanced and mesh networks. IFAC-PapersOnLine. 2020;53(2):13287-13292. doi:10.1016/j.ifacol.2020.12.159
- Ferro G, Minciardi R, Parodi L, Robba M, Rossi M. A multi-objective and multi-decision maker approach for the balancing market in distribution grids in presence of aggregators. 2020 7th International Conference on Control, Decision and Information Technologies (CoDIT). 2020. doi:10.1109/codit49905.2020.9263784
- T. R Nudell1, M. Brignone, M. Robba, A. Bonfiglio, G. Ferro, F. Delfino, and A. M. Annaswamy. Distributed Control for Polygeneration Microgrids: A Dynamic Market Mechanism Approach Control Eng Pract.(2022)
- S. Boyd, N. Parikh, E. Chu, B. Peleato, J. Eckstein, et al., “Distributed optimization and statistical learning via the alternating direction method of multipliers,” Foundations and Trends in Machine learning, vol. 3, no. 1, pp. 1–122, 2011.