2020-I Call

Application deadline: June 20-th (noon) 2020
Requirements: students enrolled at the second semester of the first year of the MSc track, with 20 CFUs and 28 GPA
Application: website
Available positions:

• 15 Computer Science and Engineering

Research topics in Computer Science and Engineering

Track CSE - Advanced Software Architectures and Methodologies

Proposers: G. Agosta Topic: Compiler Construction & Programming Language Implementation The research area covers dynamic and static compiler techniques (including decompilation) aiming at extra-functional goals such as performance, energy efficiency, security, functional and performance portability, as well as the design and implementation of parallel programming models and domain specific languages (e.g., Modelica). Compatible tracks: CSE (all subtopics), ATM (Modelica)

Proposers: D. Ardagna Topic: Machine Learning (ML) Methods for Performance Evaluation of Software Systems Nowadays, our society is supported by complex software systems often run in clouds. To obtain an efficient use of resources, a performance model is required. ML techniques are becoming popular, since are both accurate and scalable. This work involves the development of ML models for systems spanning from deep learning training on GPGPUs to Big Data. Compatible tracks: ATM, CSE, TLC

Proposers: M. M. Bersani Topic: Observability at runtime While monitorability provides evidence for predicting failures, observability aims to provide highly granular insights. The objective is to design syntactical extensions of languages and a framework that enable software observability and that allow designers to deploy observable applications. Compatible tracks: CSE

Proposers: G. Cugola Topic: Models for Event-based and Streaming Applications Many information systems need to make sense of large streams of events to detect relevant and critical situations. Our research group investigates new models and languages to "reason" on streaming data, while also coping with the volume, velocity, and variety of the events in the stream. Compatible tracks: CSE

Proposers: E. Della Valle Topic: Expressive yet Efficient Stream Reasoning Real-time decision-making is critical for many domains. From healthcare to infrastructure management, passing from smart cities. Our goal is to support reactive decision-making combining vast and noisy streams with contextual domain knowledge. To this extent, we make use of BigData Stream processing platforms like Spark, Kafka, and Flink, together with inductive and deductive reasoning algorithms. Compatible tracks: CSE

Proposers: E. Di Nitto Topic: Supporting Privacy Compliance of Microservices Applications New privacy regulations require software owners monitor and demonstrate the compliance of their applications to the agreed privacy policies. The objective is to create a framework to support: 1) definition of privacy policies, 2) privacy compliance checking in the context of microservice-based applications and 3) privacy compliance enforcement, where possible. Compatible tracks: CSE

Proposers: A. Margara Topic: Parallel and Distributed Stream Processing Processing streams of data is crucial in ICT applications, to support decision making and enable timely reactions. Our research group investigates new paradigms to make big data stream processing applications more efficient, by exploiting massively parallel hardware and distributed infrastructures. Compatible tracks: CSE

Proposers: L. Mottola Topic: The Battery-less Internet of Things Energy harvesting is redefining the energy constraints of traditional battery-powered IoT devices. However, such sources of energy are generally erratic, causing systems to shutdown unpredictably. We devise new software techniques to render IoT software immune to periods of energy unavailability. Compatible tracks: CSE, TLC

Proposers: M. Pradella Topic: Structured Context-Free Languages for Parallel Analysis and Verification The object of this work is to apply techniques based on Operator Precedence (OP) languages for: code verification through model-checking; analysis of massive structured data; runtime verification of critical applications. Compatible tracks: CSE

Proposers: M. Rossi Topic: Model-Driven development of robotic applications for service robots Robots are moving out of factories and into environments such as hospitals, homes, etc., where they cooperate with humans. The research focuses on model-driven approaches for designing Human-Robot Cooperation (HRC) applications that guarantee the safety of users and the achievement of applications’ objectives through the use of (formal) models. Compatible tracks: CSE, ATM

Track CSE - Data, Web and Society

Proposers: F. Garzotto Topic: Speech-based interactive technology This research exploits speech analysis techniques and psycholinguistic approaches to provide novel forms of conversational interaction and to enable (also with the integration of biosensors) adaptive conversational interfaces. This research has important applications in various fields, particularly for persons with special needs. Compatible tracks: BIO, CSE

Proposers: D. Martinenghi Topic: Multi-objective optimization The simultaneous optimization of different criteria is a problem naturally arising in Machine Learning, Recommender Systems, Caching, and many other scenarios involving massive amounts of data. This area targets the study of ranking algorithms and their hybridization with techniques used in the mentioned scenarios. Compatible tracks: CSE

Proposers: M. Masseroli Topic: Bioinformatics The Bioinformatics research area is focused on big life science data and information of different types increasingly available in many heterogeneous sources, and their best management, integration, mining and analysis with different machine learning techniques for knowledge discovery, to better understand complex biological phenomena. Compatible tracks: BIO, CSE

Proposers: R. M. Piro Topic: Network medicine and pathway analysis Network medicine studies molecular networks in human disease. Pathway analysis identifies processes which are altered, e.g., in cancer versus control. We will develop a new approach, based on network topology, to identify changes at different scales (small changes to a pathway-wide deregulation). Compatible tracks: BIO, CSE

Track CSE - Artificial Intelligence and Robotics

Proposers: F. Amigoni Topic: Multirobot Systems Teams of cooperative robots can provide effective and efficient solutions for performing tasks in warehouse management, information gathering, search and rescue, and patrolling. Such multirobot systems present challenges in the development of decision-theoretic planning tools to accomplish the tasks with the appropriate level of autonomy. Compatible tracks: CSE, ATM

Proposers: A. Bonarini Topic: Activity recognition in the wild from mobile robots Recognizing human activity in real environments, in real time is still a challenge, and a major issue to empower social robot applications. New, efficient AI and ML systems to support this activity should be designed to detect and classify relationships among posture, movements, objects and other people. Compatible tracks: CSE

Proposers: N. Gatti Topic: Algorithmic Game Theory, Mechanism Design, and Multiagent Learning This area, central in Artificial Intelligence, studies of the strategic interactions among agents, aiming at prescribing the best strategy for the players (Libratus), or at designing interaction mechanisms that are robust when agents are selfish and computationally efficient (kidney exchange), or at learning how to exploit the strategy of an agent that is not playing at the equilibrium. Compatible tracks: CSE, ATM

Proposers: M. Matteucci Topic: Domain adaptation for beyond visible image understanding Beyond visible imaging (e.g., multi-spectral, thermal, etc.) is widely used in earth observation and industrial inspection as well as by autonomous drones and vehicles. However few, small sized, often unlabeled, datasets are available for beyond visible data; data hungry deep models still have not been exploited. We want to investigate how domain adaptation techniques can be used for semi-supervised transfer learning from visible to beyond visible image understanding. Compatible tracks: CSE, ATM, TLC

Proposers: M. Restelli Topic: Reinforcement Learning Reinforcement learning is a learning paradigm concerned with learning to control a system so as to maximize a numerical performance measure that expresses a long-term objective. Reinforcement learning is rapidly gaining attention both from academic and industrial communities due to recent successes in a variety of sequential decision-making problems. Compatible tracks: CSE

Track CSE - System Architectures

Proposers: G. Boracchi Topic: Data-Driven Models for Non Matrix Data Machine/Deep Learning models are very successful in inference problems over matrix inputs (images). We will explore data-driven models for unconventional data (e.g. flow fields from CFD, 3D meshes from CT scans) and for comparing paired modalities (depth images vs CAD models, thermal images vs CFD) Compatible tracks: CSE

Proposers: L. Cassano Topic: Dependability Solutions for Critical Infrastructures The research area focuses on the design and development of methodologies and tools to support the design and deployment of dependable ICT for critical infrastructures. Tolerance, test and diagnosis of faults in the components of a critical infrastructure will be the target activities. Compatible tracks: CSE, ELN, ELT, TLC.

Proposers: A. Miele Topic: Design of Fault Management Strategies based on Machine Learning The project exploits machine learning techniques to design advanced fault handling strategies tailored for selected classes of applications (e.g. image processing). The goal is to reduce the area/time/power overhead of traditional redundancy-based hardening and achieve an enhanced fault classification capability. Compatible tracks: CSE, TLC, BIO.

Proposers: G. Palermo Topic: Self-Tunable Applications in High Performance and Embedded Computing Systems The research area focuses on the development of methodologies, libraries, and tools to always guarantee the efficient execution of the target code. Its importance is increasing in several computing areas requiring high level of specialization, ranging from HPC to embedded and cyber-physical systems. Compatible tracks: CSE

Proposers: C. Pilato Topic: Design Methods and Architectures for Hardware Security This area studies how to enforce the hardware security of digital circuits against IP theft and infringement, unauthorized access to sensible data, and other cyber attacks. Micro-architectural solutions and automated design methods are developed and studied based on the security threats to target. Compatible tracks: CSE

Proposers: M. Roveri Topic: Lifelong machine/deep learning for IoT and Edge The project will focus on the design of novel and theoretically-grounded Machine and Deep Learning solutions able to support the lifelong-learning paradigm in real-world time-variant technologically-constrained scenarios (such as Internet-of-Things and Edge Computing systems). Compatible tracks: CSE

Proposers: C. Silvano Topic: Architectures for Deep Neural Networks In embedded systems, the computational requirements for DCNNs along with low-power and memory constraints require lot of attention, especially considering the use of HW-accelerators. This research area investigates the design space exploration of low-power accelerators for next generation DCNNs. Compatible tracks: CSE

Track CSE - Information Systems

Proposers: C. Cappiello Topic: Context-aware data quality assessment for data analytics The thesis will focus on methods and techniques for the Data Quality assessment in smart environments, where the variety and the uncertainty of data sources are increasing. An adaptive platform able to assess and improve the quality of data on the basis of the analysis to perform will be designed and realized. Compatible tracks: CSE

Proposers: B. Pernici Topic: Information Systems - citizen science and data aggregation for SDGs Citizen Science for Monitoring Climate Impacts and Achieving Climate Resilience: the thesis will focus on the use of social media and other non-traditional data sources for more effective monitoring of indicators for Sustainable Development Goals (SDGs) by citizens. A specific focus will be on representing data at different levels of detail and methodologies and tools to collect a sufficient amount of data of good quality for statistical analyses. Compatible tracks: CSE, AUT, TLC

Proposers: M. Vitali Topic: Distributed Information Systems The research area focuses on Self-Adaptive Information Systems, exploring the synergy between information and computation. The candidate will be able to explore issues like heterogeneity of resources and the trade-off between quality of service and energy efficiency using machine learning techniques. Compatible tracks: CSE.