Data Science is an interdisciplinary field of study that has established itself in recent years in order to offer the methodological tools and technologies necessary for the management and analysis of big data and their valorisation in industry, services, and search. The phenomenon of big data has revolutionized countless sectors of economic-social activity. The phenomenon of big data has also profoundly modified the research methodologies and the development of technological innovation in numerous disciplines and applications. The main objective of this PhD is the realization of interdisciplinary research projects of Data Science that lead to the development of innovative methodologies and technologies based on the use of big data in the following fields of application:
i) Advanced digital platforms,
ii) Management of urban spaces and environmental resources
iii) Medicine and health
iv) Economic and Social Analysis
Data Science receives the decisive contribution of computer science, statistics, engineering, applied mathematics, and academic disciplines that help to understand the impact of big data in applications. These skills are widely represented in the Scientific Disciplinary Sectors that compose the PhD Committee, both with reference to the core competences of Data Science and to the applications indicated above.
I) For competences related to statistics, applied mathematics, and social-economic applications, there are components of the following SSDs
SECS-S/01 - STATISTICS.
The interest is in the scientific activity, teaching and training concerning data analysis, design and implementation of surveys and experiments in various applied areas for descriptive, interpretative and decision-making purposes. It includes the theoretical and applied developments of descriptive, exploratory, inferential and decision-making statistics in their different articulations such as mathematical statistics, planning and analysis of surveys, sampling theory, experiment plan, analysis of multivariate data, time and space series analysis, reliability and statistical quality control, biostatistics, medical statistics and environmental statistics. Computational statistics and statistical software development, modern big data management and computing issues, and application methods of both observational and experimental data are an integral part of the field.
MAT/06 - PROBABILITY AND MATHEMATICAL STATISTICS
This sector includes competences and research areas related to the study—from theoretical and applied viewpoints—of probability, the related stochastic techniques, and mathematical statistics. It also studies the stochastic aspects of reliability, queueing, decision, and game theory. These areas include, among others, the following topics: stochastic analysis of algorithms, mathematical theory of games and study of learning algorithms to compute their equilibria, study of optimal decisions under uncertainty in static and dynamic frameworks, analysis of stochastic networks and queues, study of information transmission on networks, development of mathematical tools for the analysis of complex and high-dimensional data.
SECS-S/06 - MATHEMATICAL METHODS IN ECONOMICS, IN FINANCE AND IN THE ACTUARIAL SCIENCES. The group of researchers is interested in the identification and development of mathematical methods and tools for the construction and the analysis of models and problems in the areas of: business management; finance; actuarial sciences; individual, strategic and collective choices; market analysis; risk management and, more generally, in any area of the economic and social sciences. These are complex realities, whose reading can only be made through careful collection and interpretation of the observable data. The group’s interests explicitly include the research and teaching of activities regarding computational tools and data processing techniques. The role of data is so central that not only were methods and models developed to understand economic reality through data but it is increasingly important in the field to create and develop models, even theoretical ones, that are efficient from the point of view computational, so that they are able to provide fast and correct answers to phenomena of difficult analysis.
II) For the competences related to Computer Science, Information Engineering, Bioengineering and applications in the field of advanced digital platforms, there are components of the following SSDs:
INF/01 - INFORMATICS
The interest is in scientific, teaching and training activities concerning Computer Science. The core foundational content is based on logic and computational models, computability and complexity theory, design and analysis of algorithms, information theory, coding theory and cryptography, programming languages and methodologies. The scope of Computer Science involve a large range of topics including operating systems, information systems design and administration, information systems security, databases, software engineering, artificial intelligence, neural networks, information retrieval, machine learning, systems for human-computer interaction, natural language processing, computer graphics, computer networks, networks analysis and security, distributed systems. The focus is on methodology for problem solving, with special emphasis on efficiency, reliability, and security.
ING-INF/03 - TELECOMMUNICATIONS.
The discipline is concerned with the design of future communication networks whose goal will be to enable different data services, from automated driving, virtual reality, e-health, Industry 4.0, etc, using a common platform. A large number of research avenues can find their natural environment in the Data Science Doctorate: a) Big data analytics for proactive resource allocation, which relies on the ability to predict users’ requests, which requires effective ways to extract data analytics from user and network data. b) Graph signal processing: In many applications of current interest, from computational biology to brain functionality mapping, the signals of interest reside on non-metric domains, which can be represented as graphs. Graph signal processing has to do with the theory and methods to extract relevant information from these kinds of signal.
c) Heterogenous data processing: Representation and effective processing of heterogeneous data, namely sensor data, video, mobility, positioning and other specific information, for the design of new service applications in Vehicular networks (DSRC based, LTE based, 5G) and Internet of things (LoRa based, Sigfox, NB-LTE).
ING-INF/05 - INFORMATION PROCESSING SYSTEMS
This discipline covers research and teaching activities in the field of Systems for Information processing and management. The involved scientific activities and competences address the design and implementation of information processing systems, as well as their use and management guided by engineering practices.
This area includes the theoretical foundations, methods and technologies for the principled design and implementation of realistic, effective, efficient and affordable processing of data. These three axes apply to all aspects related to information processing systems, including the large-scale programming and hardware infrastructures needed for the processing of big data and the organisation of large-scale databases, languages and techniques for Web information retrieval and data mining of document collections, knowledge engineering, machine learning, artificial intelligence and robotic systems. As a result, this area covers competences that are related to the design and implementation of computer infrastructures and information processing applications, including embedded and pervasive systems, and information systems to enable social cooperation. The contribution to the Doctoral Studies in Data Science can have even broader ramifications, through synergies with different scientific areas, including but not limited to, biology, medical sciences, statistics and economics. For example, forefront research directions include the application of advanced data representation and analysis techniques to the study of complex biological systems or the derivation of new algorithmic techniques or computational paradigms for data analysis that are latent in a number of natural complex systems and economics.
ING-INF/06 – ELECTRONICS AND INFORMATICS BIOENGINEERING
Biomedical Engineering encompasses scientific and training activities relating to: (i) basic methodologies, including processing of data and signals from biological systems, bioimaging, and representation of medical and biological knowledge; (ii) technologies, including artificial intelligence systems, health management systems and healthcare systems, information systems at different levels (patient department, hospital, region, country), medical informatics, and telemedicine; (ii) advanced research areas in biology and neuroscience, including computer techniques for biology and neurology (neuroinformatics and bioinformatics).
Incorporating these activities into the PhD course offers students the opportunity to acquire the cultural background required for a career in promising fields of medical research (4P Medicine, Network Medicine), Biology (Network Biology) and Neuroscience (Connectomics), and to acquire skills required for qualified positions in companies fostering innovation in the fields of clinical informatics, imaging informatics, health informatics, bioinformatics.
III) For the competences related to Medical Sciences and Health there are components of the following SSD:
MED/01 MEDICAL STATISTICS
The field is concerned with the scientific and didactic and educational activities as well as the possible therapeutic activity in the field of Medical Statistics of the epidemiological, biostatistical and healthcare methodologies applied to the clinic, public health and evidence based medicine (EBM), and involves the design, analysis and evaluation of both experimental and observational studies in medicine, biology, veterinary medicine, pharmacology, genetics and genomics, identification of risk factors and evaluation of health policies, impact of interventions and analysis of bio-banks and socio-sanitary databases and factors of environmental health accidents.
MED/09 - INTERNAL MEDICINE
The field includes competences and research areas related to the study and promotion, both theoretical and applied, of improving the quality of care of subjects suffering from chronic diseases such as diabetes and cardiovascular disease (complex data and large ones). In this area, among others, the following topics are included: i.) Study and creation of "dynamic" learning algorithms for predicting the patient's clinical outcomes based on past clinical data and information entered by the patient through "continuous / periodic" transmission of informations; ii.) Study and creation of "dynamic" diagnostic and therapeutic management algorithms for subjects with chronic diseases through past clinical data and informations entered by the patient through the "continuous / periodic" transmission of informations; iii.) study and creation of applications to improve patient’s lifestyle, adherence to therapy and health monitoring.
MED/26 - NEUROLOGY
Over the last decade, an increasing number of authors have used advanced wireless technologies including wearable sensors based on Inertial Measurement Units (IMU) for objective long-term monitoring of specific symptoms in patients with neurological disorders including Parkinson's disease (PD). Wireless technologies will likely provide the future framework for collection, analysis and visualization of large-scale clinical data crucially helpful in the long-term clinical management of patients with neurological disorders. The present field includes competences and research areas - from theoretical and applied viewpoints - aimed at improving the objective diagnosis, the long-term monitoring of specific motor and non-motor symptoms, and finally the overall clinical management of several neurological disorders including PD. More in detail, the field will promote studies with wireless technologies and advanced algorithms designed to automatically detect specific patterns of physiological and pathological movements, through continuous long-term monitoring, and then predict patient's clinical outcomes based on previously collected clinical data. This approach will also provide new relevant information helpful in the optimization and tailoring of pharmacological and non-pharmacological treatments of patients with PD and other neurological disorders.
MED/28 - ODONTOSTOMATOLOGIC DISEASES
The disciplinary field includes the scientific, educand training activities within the clinical data analysis and research, with specific focus on the application of network science to Dentistry and Orthodontics. Modern statistical techniques will be developed in order to combine the experimental data with the previous knowledge, available in literature or previous tests, so as to assess the probability and the presence of specific hypothesis - a highly effective method in biological and medical research.A statistical/computational approach to Dentistry and Orthodontics can be described as a dynamic biological system of interacting elements within a simplified representation that detects the interaction itself; this can provide an insight on how these elements affect (and possibly are the cause) one another. The development and in-depth study of complex networks and their application to Dentistry and Orthodontics, will explain clearly the interaction among the relevant variables, demonstrating how these variables interact with complex networks.
MED/36 - DIAGNOSTICS FOR IMAGES, RADIOTHERAPY AND NEURORADIOLOGY
The sector includes scientific and didactic activities, as well as assistance in the field of Diagnostic imaging and radiotherapy and interventional radiology, organs and systems and nuclear medicine; specific competencies include general and oncological radiation therapy and clinical radiology anatomy. It also studies radiation protection, medical radiobiology and diagnostics for images of sports activities. The sector is also focused on Neuroradiology with specific expertise in clinical neuroradiological anatomy, in general and interventional neuroradiology and in diagnostic imaging of the nervous system. It also deals with image management and analysis of archived data for scientific and epidemiological purposes. In the specific field of the PhD in Data Science, we also intend to develop advanced machine learning and deep learning techniques for radiological diagnostics.
IV) For the competences related to the Environmental and Territory Sciences there are components of the following SSD:
ICAR/06 TOPOGRAPHY AND CARTOGRAPHY.
The discipline is concerned with scientific and teaching activities related to physical and space geodesy, positioning, navigation, image processing for metric and thematic purposes (photogrammetry, remote sensing), geographic information systems (GIS), modeling and statistical analysis of geospatial data. The general topics concerns the acquisition, modeling, processing, analysis, management and sharing of geospatial data characterized by appropriate quality indices.
The growing availability of new sensors for geospatial data acquisition and high performance computing platforms for their management, analysis and sharing has extended the discipline interest to Geo Big Data and developing methodologies and models for the full exploitation of their peculiarities (volume, variety, velocity and veracity) in the context of environmental, urban, sociological and emergency management.
Methodological and application relevant topics concern the acquisition, modeling, processing and analysis of positioning and navigation data from smartphones, the development of desktop, web and mobile GIS tools for geodata science (i.e. Free and Open Source Software, Google Earth Engine), mobility data analysis and modeling, global high resolution datasets (i.e. OpenStreetMap, High Resolution Global Land Cover), Citizen science, Geo-crowdsourced and Volunteered Geographic Information (VGI) data analysis, geodata analysis for Climate Change and Land Use/Land Cover Change issues.