Corso di laurea: Data Science Programmazione per l'A.A. 2019/2020

 

 

Lo studente espliciterà le proprie scelte al momento della presentazione, tramite INFOSTUD, del piano di completamento o del piano di studio individuale, secondo quanto stabilito dal regolamento didattico del corso di studio.

Primo anno

Primo semestre

Insegnamento	CFU	SSD	Ore Lezione	Ore Eserc.	Ore Lab	Ore Studio	Attività	Lingua
1047221 - ALGORITHMIC METHODS OF DATA MINING AND LABORATORY (obiettivi) ○ The course presents the main algorithmic techniques of data mining, necessary for data science. They offer to the student the basis for analyzing data for a variety of applications that deal with semistructured or unstructured data, such as textual data, transactions, and graph and information-network data. At the end of the course the student will have a knowledge of the main theoretical ideas of data mining, as well as some basic knowledge and experience in using programming tools for analyzing and mining data. - ANAGNOSTOPOULOS ARISTIDIS (programma) - Text mining - Basic algorithms for mining - Basic structures for data mining - Clustering - Graph mining   J. Leskovec, A. Rajaraman, and J. Ullman, "Mining of Massive Datasets," Cambridge University Press C. Aggarwal, "Data Mining: The Textbook," Springer (must be downloaded from Sapienza) M. J. Zaki and W. Meira, Jr., "Data Mining and Analysis: Fundamental Concepts and Algorithms," Cambridge University Press R. Zafarani, M. A. Abbasi, and H. Liu, "Social Media Mining: An Introduction," Cambridge University Press C. D. Manning, P. Raghavan and H. Schütze, "Introduction to Information Retrieval," Cambridge University Press (Date degli appelli d'esame) - CHATZIGIANNAKIS IOANNIS (programma) - Text mining - Basic algorithms for mining - Basic structures for data mining - Clustering - Graph mining   J. Leskovec, A. Rajaraman, and J. Ullman, "Mining of Massive Datasets," Cambridge University Press C. Aggarwal, "Data Mining: The Textbook," Springer (must be downloaded from Sapienza) M. J. Zaki and W. Meira, Jr., "Data Mining and Analysis: Fundamental Concepts and Algorithms," Cambridge University Press R. Zafarani, M. A. Abbasi, and H. Liu, "Social Media Mining: An Introduction," Cambridge University Press C. D. Manning, P. Raghavan and H. Schütze, "Introduction to Information Retrieval," Cambridge University Press	9	ING-INF/05	32	-	60	-	Attività formative caratterizzanti	ENG
1047264 - FUNDAMENTALS OF DATA SCIENCE AND LABORATORY (obiettivi) Learning from data in order to gain useful predictions and insights. At the end of the course students will have an understanding of the basic programming skills needed for data analysis and visualization. They will also have familiarity of the typical data processing workflow of data preparation and scraping, visualization and exploratory analysis and final statistical modeling. Students will become familiar with the main Python libraries for data science. - GALASSO FABIO (programma) This course is an introduction to the basic toolkit of data analysis using the R programming language, and an introduction to programming using the Python programming language. The main data analysis topics covered are the following (not necessarily in chronological order): regression; classification; computational complexity and NP-hardness; clustering; linear and integer programming.   - Bertsimas, O'Hair, Pulleyblank: "The Analytics Edge". Dynamic Ideas LLC, 2016. ISBN: 098991089X, 9780989910897. - EdX Course: https://www.edx.org/course/analytics-edge-mitx-15-071x-3 R: - R for data science: http://r4ds.had.co.nz Python: - The Python Tutorial: https://docs.python.org/3/tutorial/ (Date degli appelli d'esame)	9	INF/01	32	-	60	-	Attività formative caratterizzanti	ENG
10589600 - STATISTICAL METHODS IN DATA SCIENCE AND LABORATORY (obiettivi) Learning goals Statistical Methods in Data Science is a two-semester course aimed at providing the fundamental tools for: setting up probabilistic models; understanding the basic principles of the main inferential problems: estimation, hypothesis testing, model checking and forecasting; understanding and contrasting the two main inferential paradigms, namely frequentist and Bayesian statistics; implementing inference on observed data through both optimization and simulation-based (approximation) techniques such as: Bootstrap Monte Carlo Monte Carlo Markov Chain (MCMC) understanding comparative merits of alternative strategies developing statistical computations within a suitable software environment like R (www.r-project.org), OpenBUGS (http://openbugs.net/w/FrontPage) and STAN (http://mc-stan.org/). Knowledge and understanding On successful completion of this course, students will: know the main statistical principles, inferential problems, paradigms and algorithms; assess the empirical and theoretical performance of different modeling approaches; know the main platforms, programming languages to develop effective implementations. Applying knowledge and understanding Besides the understanding of theoretical aspects, thanks to applied homeworks and a dedicated laboratory in the second semester focused on Bayesian modeling, students will be constantly challenged to use and evaluate all the techniques they have learned as well as to propose new modelization suitable for specific tasks at hand. Making judgements On successful completion of this course, students will develop a positive critical attitude towards the empirical and theoretical evaluation of statistical methodologies and results. Communication skills In preparing the report and oral presentation for the final project of the second semester laboratory, students will learn how to effectively communicate information, ideas, problems and solutions to specialists but also to a general audience. Learning skills In this course the students will develop the skills necessary for a successful understanding and application of new statistical methodologies together with their effective implementation. The goal is of course to grow an active attitude towards continued learning throughout a professional career.
- STATISTICAL METHODS IN DATA SCIENCE AND LABORATORY I (obiettivi) Learning goals Statistical Methods in Data Science is a two-semester course aimed at providing the fundamental tools for: setting up probabilistic models; understanding the basic principles of the main inferential problems: estimation, hypothesis testing, model checking and forecasting; understanding and contrasting the two main inferential paradigms, namely frequentist and Bayesian statistics; implementing inference on observed data through both optimization and simulation-based (approximation) techniques such as: Bootstrap Monte Carlo Monte Carlo Markov Chain (MCMC) understanding comparative merits of alternative strategies developing statistical computations within a suitable software environment like R (www.r-project.org), OpenBUGS (http://openbugs.net/w/FrontPage) and STAN (http://mc-stan.org/). Knowledge and understanding On successful completion of this course, students will: know the main statistical principles, inferential problems, paradigms and algorithms; assess the empirical and theoretical performance of different modeling approaches; know the main platforms, programming languages to develop effective implementations. Applying knowledge and understanding Besides the understanding of theoretical aspects, thanks to applied homeworks and a dedicated laboratory in the second semester focused on Bayesian modeling, students will be constantly challenged to use and evaluate all the techniques they have learned as well as to propose new modelization suitable for specific tasks at hand. Making judgements On successful completion of this course, students will develop a positive critical attitude towards the empirical and theoretical evaluation of statistical methodologies and results. Communication skills In preparing the report and oral presentation for the final project of the second semester laboratory, students will learn how to effectively communicate information, ideas, problems and solutions to specialists but also to a general audience. Learning skills In this course the students will develop the skills necessary for a successful understanding and application of new statistical methodologies together with their effective implementation. The goal is of course to grow an active attitude towards continued learning throughout a professional career. - BRUTTI PIERPAOLO (programma) Part I Probability --------------------- Outcomes, Events & Probability. Conditional Probability & Independence. Bayes' Theorem: Interpretation & Use. Random Variables and Random Vectors. Directed Graphical Models. Expected Value, Variance and Covariance. Univariate and Multivariate Distributions. Generating Functions and Convergence Theorems. Part II Statistics --------------------- The Empirical Distribution and the Bootstrap Statistical Modelling: the Likelihood Function. Parameter Estimation: Point and Interval Estimation. Hypothesis Testing. Alternative Inferential Frameworks: Frequentist vs Bayesian Inference.   Riferimenti principali: F.M. Dekking, C. Kraaikamp, H.P. Lopuhaa, L.E. Meester (2007). A Modern Introduction to Probability and Statistics. Springer. L. Wasserman (2004). All of Statistics: A Concise Course in Statistical Inference. Springer.	9	SECS-S/01	36	54	-	-	Attività formative caratterizzanti	ENG

Secondo semestre

Insegnamento	CFU	SSD	Ore Lezione	Ore Eserc.	Ore Lab	Ore Studio	Attività	Lingua
10589600 - STATISTICAL METHODS IN DATA SCIENCE AND LABORATORY (obiettivi) Learning goals Statistical Methods in Data Science is a two-semester course aimed at providing the fundamental tools for: setting up probabilistic models; understanding the basic principles of the main inferential problems: estimation, hypothesis testing, model checking and forecasting; understanding and contrasting the two main inferential paradigms, namely frequentist and Bayesian statistics; implementing inference on observed data through both optimization and simulation-based (approximation) techniques such as: Bootstrap Monte Carlo Monte Carlo Markov Chain (MCMC) understanding comparative merits of alternative strategies developing statistical computations within a suitable software environment like R (www.r-project.org), OpenBUGS (http://openbugs.net/w/FrontPage) and STAN (http://mc-stan.org/). Knowledge and understanding On successful completion of this course, students will: know the main statistical principles, inferential problems, paradigms and algorithms; assess the empirical and theoretical performance of different modeling approaches; know the main platforms, programming languages to develop effective implementations. Applying knowledge and understanding Besides the understanding of theoretical aspects, thanks to applied homeworks and a dedicated laboratory in the second semester focused on Bayesian modeling, students will be constantly challenged to use and evaluate all the techniques they have learned as well as to propose new modelization suitable for specific tasks at hand. Making judgements On successful completion of this course, students will develop a positive critical attitude towards the empirical and theoretical evaluation of statistical methodologies and results. Communication skills In preparing the report and oral presentation for the final project of the second semester laboratory, students will learn how to effectively communicate information, ideas, problems and solutions to specialists but also to a general audience. Learning skills In this course the students will develop the skills necessary for a successful understanding and application of new statistical methodologies together with their effective implementation. The goal is of course to grow an active attitude towards continued learning throughout a professional career.
- STATISTICAL METHODS IN DATA SCIENCE AND LABORATORY II (obiettivi) Learning goals Statistical Methods in Data Science is a two-semester course aimed at providing the fundamental tools for: setting up probabilistic models; understanding the basic principles of the main inferential problems: estimation, hypothesis testing, model checking and forecasting; understanding and contrasting the two main inferential paradigms, namely frequentist and Bayesian statistics; implementing inference on observed data through both optimization and simulation-based (approximation) techniques such as: Bootstrap Monte Carlo Monte Carlo Markov Chain (MCMC) understanding comparative merits of alternative strategies developing statistical computations within a suitable software environment like R (www.r-project.org), OpenBUGS (http://openbugs.net/w/FrontPage) and STAN (http://mc-stan.org/). Knowledge and understanding On successful completion of this course, students will: know the main statistical principles, inferential problems, paradigms and algorithms; assess the empirical and theoretical performance of different modeling approaches; know the main platforms, programming languages to develop effective implementations. Applying knowledge and understanding Besides the understanding of theoretical aspects, thanks to applied homeworks and a dedicated laboratory in the second semester focused on Bayesian modeling, students will be constantly challenged to use and evaluate all the techniques they have learned as well as to propose new modelization suitable for specific tasks at hand. Making judgements On successful completion of this course, students will develop a positive critical attitude towards the empirical and theoretical evaluation of statistical methodologies and results. Communication skills In preparing the report and oral presentation for the final project of the second semester laboratory, students will learn how to effectively communicate information, ideas, problems and solutions to specialists but also to a general audience. Learning skills In this course the students will develop the skills necessary for a successful understanding and application of new statistical methodologies together with their effective implementation. The goal is of course to grow an active attitude towards continued learning throughout a professional career. - TARDELLA LUCA   Teaching material will be also delivered through the Sapienza elearning platform Moodle at the following address: https://elearning2.uniroma1.it/course/view.php?id=4951 Course Material and Main Reference Books • Course lecture notes (slide available at the https://elearning2.uniroma1.it/course/view.php?id=7253) • R or R+Jags commented codes • Peter Hoff, A First Course in Bayesian Statistical Methods. Springer-Verlag Inc, 2009. • Jean-Michel Marin and Christian P. Robert, Bayesian Core: A Practical Approach to Computational Bayesian Statistics, Springer, 2007 • Christian P. Robert and George Casella. Monte Carlo statistical methods (2nd ed.. Springer-Verlag Inc, 2004. • Ioannis Ntzoufras, Bayesian Modeling Using WinBUGS. Wiley, 2009. • Peter Congdon. Bayesian Statistical Modelling (2nd ed.). Wiley, 2006 (Date degli appelli d'esame)	3	SECS-S/01	12	18	-	-	Attività formative caratterizzanti	ENG
Gruppo opzionale: GRUPPO OPZIONALE A - (visualizza) 6                1047215 - INTELLECTUAL PROPERTY COMPETITION AND DATA PROTECTION LAW (obiettivi) The aim of the course is to provide students with an overview of the functioning of intellectual property, competition and data protection law from both an economic and legal perspective. By the end of the course students are expected to have acquired a general understanding of the main policy issues involved, and should be able to identify and apply the relevant legal rules, both substantial and procedural, in situations that can be considered routinary to professionals and businesses operating in the data science industry. - BOSO CARETTA ALESSANDRO (Date degli appelli d'esame) 6  IUS/04  24  36  -  -  Attività formative caratterizzanti  ENG 1047212 - ECONOMICS OF NETWORK INDUSTRIES (obiettivi) Knowledge and understanding The aim of the course is to introduce students to the new information economy and the economics of network industries. Students are expected to gain insight into how the specific features of technology and demand affect market structure, firms’ strategies and business models, as well as public policy in network industries. Applying knowledge and understanding By the end of the course, students should be able to use methods and models of microeconomics and industrial organization to understand and analyze the competitive dynamics in the new information economy, and specifically in network industries. Making judgements Lectures, practical exercises and problem-solving sessions will provide students with the ability to assess the main strengths and weaknesses of theoretical models when used to explain empirical evidence and case studies in the new information economy. Communication By the end of the course, students are able to point out the main features of the new information economy and network industries, and to discuss relevant information, ideas, problems and solutions both with a specialized and a non-specialized audience. These capabilities are tested and evaluated in the final written exam and possibly in the oral exam as well as in the project work. Lifelong learning skills Students are expected to develop those learning skills necessary to undertake additional studies on relevant topics in the field of the new information economy with a high degree of autonomy. During the course, students are encouraged to investigate further any topics of major interest, by consulting supplementary academic publications, specialized books, and internet sites. These capabilities are tested and evaluated in the final written exam and possibly in the oral exam as well as in the project work, where students may have to discuss and solve some new problems based on the topics and material covered in class. - REVERBERI PIERFRANCESCO (programma) 1. Technology, demand and market structure in network industries • Economies of scale and scope; Sunk costs • Complementary goods or services; Network effects • Market concentration Readings: - Shapiro-Varian, Chapter 1 - Slides 2. Market power and pricing strategies • Personalized pricing and group pricing • Menu pricing • Intertemporal pricing • Bundling • Switching costs and lock-in Readings: - Shapiro-Varian, Chapters 2, 3, 5, 6 - Church J., R. Ware (2000), Industrial organization: A strategic approach, McGraw-Hill (available at http://works.bepress.com/jeffrey_church/23): Chapter 2, sections 2.3 (2.3.1), 2.4 (2.4.1, 2.4.2), 2.5; Chapter 4, sections 4.1 (4.1.1, 4.1.2), 4.3 (4.3.1), 4.5; Chapter 5 (except for section 5.4.5); Chapter 25 (section 25.4) - Klemperer P. (1995), Competition when consumers have switching costs: An overview, Review of Economic Studies 62, pp. 515-539: Sections 1, 2, 3, 7, 8 - Kobayashi B. (2005), Does economics provide a reliable guide to regulating commodity bundling by firms? A survey of the economic literature, Journal of Competition Law and Economics 1, pp. 707-746: Sections 1, 2 (A, C, D), 3 (B), 4 (B) 3. Network externalities, standardization, and compatibility • Demand side: users’ coordination and technology adoption • Supply side: Compatibility strategies vs. standards wars • Public policy in network industries Readings: - Shapiro-Varian, Chapters 7, 8, 9, 10 - Shy O. (2011), A short survey of network economics, Review of Industrial Organization 38, pp. 119-149: Sections 1, 2, 3 (3.1, 3.2, 3.4), 4, 6, 8 4. Two-sided markets • Key features and examples • Pricing issues and implications for public policy Readings: - Armstrong M. (2006), Competition in two-sided markets, Rand Journal of Economics 37, pp. 668-691: Sections 1, 2, 3 - Evans D.S., R. Schmalensee (2007), The industrial organization of markets with two-sided platforms, Competition Policy International 3, pp. 151-179: Sections 1, 2, 3, 4, 7 5. Telecommunications (brief overview) • Natural monopoly and regulation • Access pricing and competition Readings: - Church J., R. Ware (2000), Industrial organization: A strategic approach, McGraw-Hill (available at http://works.bepress.com/jeffrey_church/23): Chapters 24, 25, 26 - overview • Further readings: Collection of relevant articles from the press (file ‘Stories’ on the web site)   The main reference textbook is: · Shapiro C., H. Varian (1999), Information Rules: A Strategic Guide to the Network Economy, Harvard Business Press. (Date degli appelli d'esame) 6  SECS-P/06  24  36  -  -  Attività formative caratterizzanti  ENG
1047223 - NETWORKING FOR BIG DATA AND LABORATORY (obiettivi) Obiettivi generali: Lo scopo del corso è fornire agli studenti gli strumento per comprendere I principi del networking e le principali tecnologie di rete. Il corso è focalizzato sull’evoluzione della rete Internet per il supporto dei big data e del cloud, con particolare attenzione alle soluzioni di rete per i data centers. La prima parte del corso sarà necessaria per rendere omogeneo il livello della classe e per definire i concetti e i termini tecnici di base. Il corso prevede anche l’utilizzo di un emulatore di rete e di un analizzatore di traffico per lo svolgimento di attività pratiche di laboratorio. Obiettivi specifici: Conoscenza e capacità di comprensione: conoscere i principali protocolli di rete per la realizzazione di una rete IP. Conoscenza e capacità di comprensione applicate: saper applicare i principi del networking per realizzare una rete emulata funzionante e per analizzare in maniera critica il traffico all’interno di una rete Autonomia di giudizio: capacità di individuare criticamente gli elementi di debolezza delle soluzioni architetturali studiate nello scenario di un data center per la gestione dei big data Capacità di apprendere: capacità di proseguire gli studi successivi riguardanti tematiche avanzate di networking. - CIANFRANI ANTONIO (programma) Modulo 1. FONDAMENTI DI NETWORKING 1a. Pila protocollare TCP/IP 1b. Strato di trasporto: TCP e UDP 1c. Strato IP: indirizzamento e instradamento 1d. Strato di collegamento: indirizzi fisici e MAC forwarding 1e. Attività di laboratorio: emulatore di rete IP (Netkit) and analizzatore di traffico (Wireshark) Module 2. NETWORKING AVANZATO 2a. Virtual LAN (VLAN) 2b. Virtual eXtensible LAN (VXLAN) 2c. Locator/Identifier Separation Protocol (LISP) 2d. Software Defined Networking (SDN) 2e. Network Function Virtualization (NFV) 2f. Protocollo HTTP Module 3. DATA CENTER 3a. Implicazioni del cloud computing. 3b. Data centers: architetture, topologie, indirizzamento, instradamento, caratteristiche del traffico, affidabilità. 3c. Gestione delle tisorse e scheduling. 3d. Ottimizzazione del consumo energetico. 3e. Controllo di congestione (QCN).   - James F. Kurose and Keith Ross, “Computer Networking: A Top-Down Approach Featuring the Internet”, Pearson - Lucidi delle lezioni (Date degli appelli d'esame) - BAIOCCHI ANDREA (programma) Descrizione ad alto livello del ‘cloud computing’. Reti per i ‘data center’. Struttura e componenti. Progetto topologico dei ‘data center’. Indirizzamento e instradamento: confronto tra soluzioni a livello IP e Ethernet. Alberi ricoprenti e algoritmo di ‘Spanning tree’ in Ethernet. Caratteristiche del traffico nei ‘data center’. Affidabilità. Gestione e assegnazione delle risorse. Controllo della congestione: Quantized Congestion Notification (QCN) nelle reti Ethernet; Data Center TCP (TCP). Consumi energetici nei ‘data center’.   1. Yang Liu et alii, Data Center Networks Topologies Architectures and Fault Tolerance Characteristics, Springer, 2013. General overview – main text on the Data Center networking. 2. N. da Fonseca and R. Boutaba (Eds.), Cloud services, networking, and management, IEEE press, 2015. Chapters 1 (Cloud Computing), 4 (Data Center networks), 8 (Energy management), 10 (Resource scheduling), 12 (Survivability). 3. D. Bertsekas, and R. Gallager, Data networks, Prentice Hall, 1992. Chapter 5, Secs. 5.2.1, 5.2.2, p. 387-393: Spanning tree and minimum spanning tree (MST): general definitions and algorithms. 4. W. Stallings, Local and metropolitan area networks, Prentice Hall, 1997. Chapter 14, Secs. 14.1, 14.2, 14.3, p. 482-499: Spanning tree protocol. 5. Samer Salam and Ali Sajassi, Provider Backbone Bridging and MPLS: Complementary Technologies for Next-Generation Carrier Ethernet Transport, IEEE Communication Magazine, March 2008, pages 77-63. Covers the specific topic of Provider Backbone Bridges in the relevant Section of the paper. 6. Mohammad Alizadeh et alii, Data Center Transport Mechanisms: Congestion Control Theory and IEEE Standardization, Forty-Sixth Annual Allerton Conference Allerton House, UIUC, Illinois, USA September 23-26, 2008. Covers the specific topic of QCN in Ethernet networks (Secs. I, II). 7. Mohammad Alizadeh et alii, Data Center TCP (DCTCP), ACM SIGCOMM’10, August 30–September 3, 2010, New Delhi, India. Covers the specific topic of TCP re-design for Data Centers. SITO WEB DI RIFERIMENTO https://web.uniroma1.it/netlab/networking-big-data-and-laboratory	9	ING-INF/03	40	-	48	-	Attività formative caratterizzanti	ENG
Gruppo opzionale: GRUPPO OPZIONALE B - (visualizza) 18                1047197 - DATA MANAGEMENT FOR DATA SCIENCE (obiettivi) The main goal of the course is to present the basic concepts of data management systems. The first part of the course introduces the main aspects of relational database systems, including basic functionalities, file and index organizations, and query processing. The second part of the course aims at presenting the main non-relational approaches to data management, in particular, multidimensional data management, large-scale data management, and open data management. - ROSATI RICCARDO (programma) 1. The structure of a Data Base Management System - Basic functionalities of data server 2. Physical structures for data - file organizations, indexed organizations 3. Principles of query evaluation - evaluation of relational algebra operators, query planning and optimization 4. Multidimensional data - OLAP Queries, Structures for multidimensional data, OLAP query evaluation 5. Large-scale data management - Distributed query evaluation, NoSQL databases, graph databases 6. Open data management - open data, linked open data, RDF databases   Handouts provided by the teacher (Date degli appelli d'esame) - LEMBO DOMENICO (programma) 1. The structure of a Data Base Management System - Basic functionalities of data server 2. Physical structures for data - file organizations, indexed organizations 3. Principles of query evaluation - evaluation of relational algebra operators, query planning and optimization 4. Multidimensional data - OLAP Queries, Structures for multidimensional data, OLAP query evaluation 5. Large-scale data management - Distributed query evaluation, NoSQL databases, graph databases 6. Open data management - open data, linked open data, RDF databases   Handouts provided by the teacher 6  ING-INF/05  24  36  -  -  Attività formative caratterizzanti  ENG 1047205 - CLOUD COMPUTING (obiettivi) Obiettivi generali: Lo scopo del corso è quello di fornire agli studenti i concetti di base dei sistemi distribuiti e quindi di concentrarsi sulle tecnologie di cloud computing. Il corso copre aspetti teorici e pratici con un focus su esempi reali. Alla fine del corso si suppone che gli studenti siano in grado di scegliere, configurare e utilizzare i servizi cloud e progettare e distribuire architetture scalabili ed applicazioni elastiche. Obiettivi specifici: Conoscenza e capacità di comprensione: Al termine del corso, lo studente sarà in grado di descrivere e spiegare - i concetti generali dei sistemi distribuiti - il concetto di virtualizzazione di sistema e applicazione - i meccanismi e gli algoritmi utilizzati nel cloud computing - le tecnologie per lo storage cloud - i framework per l’elaborazione dei bigdata - i problemi di sicurezza informatica e le soluzioni nel cloud computing Conoscenza e capacità di comprensione applicate: Al termine del corso, lo studente sarà in grado di - progettare e implementare un'architettura scalabile e distribuire un'applicazione elastica - presentare risultati pratici sotto forma di rapporto tecnico - analizzare e presentare lavori scientifici - selezionare, configurare ed eseguire servizi cloud utilizzando la GUI e l'API di gestione offerte dai provider IaaS - progettare e configurare infrastrutture scalabili e applicazioni distribuite elastiche. - fare scelte progettuali che tengano conto dei problemi di sicurezza informatica Autonomia di giudizio: Al termine del corso, lo studente: - sara’ in grado di valutare e confrontare le tecnologie cloud e i servizi cloud, nonché i framework di elaborazione dei big data - sara’ in grado di identificare, valutare e confrontare soluzioni all'avanguardia - rafforzera’ la sua capacità di pensiero critico Abilità comunicative: Al termine del corso lo studente: - sara’ in grado di discutere e trasmettere la propria opinione sulle tecnologie cloud - sara’ grado di presentare l'analisi di un argomento selezionato a un vasto pubblico Capacità di apprendere: Durante il corso, lo studente svilupperà e migliorerà la sua capacità di pensiero critico attraverso lo studio e l'analisi di lavori scientifici e di documentazione tecnica. Inoltre, lo studente migliorerà la sua capacità di integrare informazioni da diverse fonti, ad es. libri, documenti tecnici/scientifici ed esperienze pratiche. - CASALICCHIO EMILIANO (programma) OVERVIEW The course will cover the fundamentals of Cloud computing: Processing, Storage of large data sets, performance, scalability, reliability and security. It also will cover the organization of popular public cloud platforms like Google and Amazon and the use of cloud application programming frameworks like MapReduce and Spark. The course will deal with theoretical and practical aspects. Specific assignments and hands-on sessions will give you the possibility to develop your skill. Examples of technologies you will learn about are: Docker, Amazon Web Services, Spark, MapReduce. ARGUMENTS - The big picture of cloud computing, definition, business drivers, benefit and risks. - Cloud computing reference architecture - Enabling technologies - Virtualization (system level and process level) - Storage (cloud file systems, storage serving) - Monitoring - Adaptation and automation - Scalability and high availability - Security - Big data programming frameworks For the latest version of the program and more information check the course web site https://sites.google.com/di.uniroma1.it/cloudcomputingcourse/   - Mastering Cloud Computing, Rajkumar Buyya; Christian Vecchiola; S.Thamarai Selvi Morgan Kaufmann, ISBN: 978-0-12-409539-7 - Cloud Computing: Concepts, Technology & Architecture, Thomas Erl, Zaigham Mahmood, Ricardo Puttini, Prentice Hall/PearsonPTR ISBN: 978-0-13-338752-0 The learning material will be integrated with research papers and technical documentation (Date degli appelli d'esame) 6  INF/01  24  36  -  -  Attività formative caratterizzanti  ENG 1047200 - DATA MINING TECHNOLOGY FOR BUSINESS AND SOCIETY (obiettivi) The course will present fundamental technologies for advanced data mining applications. The course will start with presenting the methodologies for storing and retrieving information on the Web, mining application logs, mining social media, collaborative filtering and personalization. The course will also present the basic technlogies for mining geo-spatial data and data produced from large-scale sensing environments. Applications will include mining of consumer preferences, computational advertising, online marketplaces, digital marketing and mining data in physical spaces. As part of the course students will carry on a field study on a relevant use case for a selected application. - LEONARDI STEFANO (programma) ○ Basic concepts of Web Information Retrieval and Search Engine Architecture. ○ Web Usage and Social Media Data Mining ○ Recommendation Systems and Personalization ○ Data Mining in Computational Advertising and Online Markets. ○ Digital marketing and customer preferences. ○ Online labour marketplaces and Collaborative systems. ○ Geo-spatial Data mining ○ Mining sensing environments   References Ref 1. [[http://nlp.stanford.edu/IR-book/information-retrieval-book.html Christopher D. Manning, Prabhakar Raghavan, harga hp Hinrich Schueze Introduction to Information Retrieval, Cambridge University Press, 2008 ]] Ref 2. J. Leskovec, A. Rajaraman, and J. Ullman, Mining of Massive Datasets, Cambridge University Press. A textbook reference on algorithm design for Master students: Ref 3. J. Kleinberg and E. Tardos. Algorithm Design. Pearson Education (Paperback), 2013. (Date degli appelli d'esame) 6  ING-INF/05  24  36  -  -  Attività formative caratterizzanti  ENG
Gruppo opzionale: GRUPPO OPZIONALE C - (visualizza) 6                1047208 - STATISTICAL LEARNING (obiettivi) Devising new machine learning methods and statistical models is a fun and extremely fruitful “art”. But these powerful tools are not useful unless we understand when they work, and when they fail. The main goal of statistical learning theory is thus to study, in a statistical framework, the properties of learning algorithms mainly in the form of so-called error bounds. This course introduces the techniques that are used to obtain such results, combining methodology with theoretical foundations and computational aspects. It treats both the basic principles to design successful learning algorithms and the “science” of analyzing an algorithm’s statistical properties and performance guarantees. Theorems are presented together with practical aspects of methodology and intuition to help students develop tools for selecting appropriate methods and approaches to problems in their own data analyses. Methods for a wide variety of applied problems will be explored and implemented on open-source software like R (www.r-project.org), Keras (https://keras.io/) and TensorFlow (https://www.tensorflow.org/). Knowledge and understanding On successful completion of this course, students will: know the main learning methodologies and paradigms with their strengths and weakness; be able to identify a proper learning model for a given problem; assess the empirical and theoretical performance of different learning models; know the main platforms, programming languages and solutions to develop effective implementations. Applying knowledge and understanding Besides the understanding of theoretical aspects, thanks to applied homeworks and a final project possibly linked to hackathons or other data analysis competitions, the students will constantly be challenged to use and evaluate modern learning techniques and algorithms. Making judgements On successful completion of this course, students will develop a positive critical attitude towards the empirical and theoretical evaluation of statistical learning paradigms and techniques. Communication skills In preparing the report and oral presentation for the final project, students will learn how to effectively communicate original ideas, experimental results and the principles behind advanced data analytic techniques in written and oral form. They will also understand how to offer constructive critiques on the presentations of their peers. Learning skills In this course the students will develop the skills necessary for a successful understanding as well as development of new learning methodologies together with their effective implementation. The goal is of course to grow a active attitude towards continued learning throughout a professional career. - BRUTTI PIERPAOLO (programma) 1. Review of basic probability and inference. 2. Concentration of measure. 3. Basics of convex optimization. 4. Statistical functional: bootstrap & subsampling. 5. Nonparametric Regression and Density estimation: kernels and RKHS. 6. Nonparametric Classification. 7. Nonparametric Clustering: k-means, density clustering. 8. Graphical Models and their applications: parametric and nonparametric approaches. 9. Hints of Nonparametric Bayes 10. Minimaxity & Sparsity Theory.   Riferimenti principali Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani (2013). An Introduction to Statistical Learning with Applications in R. Available at: http://www-bcf.usc.edu/~gareth/ISL/ Larry Wasserman (2005). All of Nonparametric Statistics. Stephen Boyd and Lieven Vandenberghe. Convex Optimization. Available at: http://stanford.edu/~boyd/cvxbook/ (Date degli appelli d'esame) 6  SECS-S/01  24  36  -  -  Attività formative caratterizzanti  ENG 1047209 - QUANTITATIVE MODELS FOR ECONOMIC ANALYSIS AND MANAGEMENT (obiettivi) General Objectives of the course The general objectives of the course are: - Present a general framework for the development of quantitative models for economic analysis and management; - Provide the basic concepts and a guide to analyse the specialised literature; - Propose a unified framework on the main methodologies available to compare the productivity and efficiency of Decision Making Units (DMUs); - Introduce to the relevant roles played by the data for the development of effective quantitative models of socio-economic systems; - Make an introduction to the main softwares available to implement the quantitative models presented during the course; - Provide laboratory sessions to implement the quantitative models presented during the course in practice; - Present several applications in the field of economics and management, including public sector services as potential group project works, to be developed by the students according to their personal interest and background; - Interact with students through seminars, assisted laboratory, oral presentations and the realization of a project work on real data. Specific objectives of the course • KNOWLEDGE AND UNDERSTANDING: DEMONSTRATE THE KNOWLEDGE OF THE BASIC METHODS FOR THE DEVELOPMENT OF QUANTITATIVE MODELS FOR ECONOMIC ANALYSIS AND MANAGEMENT ; • ABILITY TO APPLY KNOWLEDGE AND UNDERSTANDING: TO BE ABLE TO DEVELOP QUANTITATIVE ECONOMIC MODELS ON THE BASE OF THE KNOWLEDGE AND TECHNIQUES LEARNED DURING THE COURSE; • JUDGMENT AUTONOMY: TO BE ABLE TO DEVELOP A QUANTITATIVE ECONOMIC MODEL WITH CRITICAL SPIRIT, CHOOSING THE APPROPRIATE METHOD AND CORRECTLY IMPLEMENTING IT. • COMMUNICATION SKILLS: BEING ABLE TO COMMUNICATE THE RESULTS OF THE ANALYSIS AND ITS INFORMATION TO DIFFERENT TYPES OF INTERLOCUTORS; • LEARNING SKILLS: TO DEVELOP THE NECESSARY SKILLS TO APPLY AND DEVELOP AUTONOMOUSLY THE METHODS AND MODELS LEARNED DURING THE COURSE. - DARAIO CINZIA (programma) The course is composed by the following main sections, organized in modules. -Section I. INTRODUCTION 1) A three-dimensional framework for a quantitative approach to the economic analysis and management. -Section II. DATA 2) Nature, collection, semantic modelling, and analysis of big data and little data within organizations and socio-economic systems. 3) Classification systems and information taxonomies for the management of organizations and socio-economic systems. -Section III. TOOLS Quantitative tools for economic modelling and management 4) Cost Benefit Analysis 5) Regression Methods for Continuous and Discrete Responses from Parametric to Nonparametric Approaches 6) Productivity and Efficiency Analysis: main techniques in an unified approach from parametric to nonparametric models. 7) Statistical Tools from the Physics of Complex Systems. 8) Sensitivity Analysis and Sensitivity Auditing techniques. -Section IV. APPLICATIONS 9) Available data sources for empirical analysis in economics and management 10) Applications in economics and management, including public sector services. Part one. Outline of the existing literature. 11) Applications in economics and management, including public sector services. Part two. Developments during the Project work activities (see below). Group project works The group project works will be defined according to the interest of students. The following broad projects areas will be available: 1) Estimation of socio-economic models using ISTAT (http://www.istat.it/it/prodotti/banche-dati) and EUROSTAT (http://ec.europa.eu/eurostat) data; 2) Empirical analysis of education, science and technology systems, with data coming from ongoing European research projects, including the ETER project (http://eter.joanneum.at/imdas-eter/ ); 3) Application of statistical tools from the physics of complex systems to compare the scientific performance of countries, and case studies in collaboration with the Italian Institute of Technology (https://www.iit.it/it/home.html, http://lns.iit.it/ ); 4) Applications of sensitivity analysis and sensitivity auditing, in collaboration with the Joint Research Center of the European Commission, Ispra (https://ec.europa.eu/jrc/ ).   A base of the course is the material contained in: Daraio C. (2016) Eds., Challenges of Big Data for Economic Modeling and Management: Tools from Efficiency Analysis, Sensitivity Analysis, Sensitivity Auditing and Physics of Complex Systems. Proceedings of the Workshop of the 10-11 November 2015, DIAG Sapienza University of Rome, Edizioni Efesto, Rome. During the course the Lecture Notes and additional materials (including reference lists, articles and so on) will be distributed. (Date degli appelli d'esame) 6  ING-IND/35  24  36  -  -  Attività formative caratterizzanti  ENG
- - A SCELTA DELLO STUDENTE	6		24	36	-	-	Attività formative a scelta dello studente (art.10, comma 5, lettera a)	ITA

Secondo anno

Primo semestre

Insegnamento	CFU	SSD	Ore Lezione	Ore Eserc.	Ore Lab	Ore Studio	Attività	Lingua
Gruppo opzionale: GRUPPO OPZIONALE B - (visualizza) 18                1047214 - DATA PRIVACY AND SECURITY (obiettivi) General Objectives Ensuring the privacy of personal data, and securing the computing infrastructures, are key concerns when collecting and analyzing sensitive data sets. Example of these data sets include medical data, personal communication, personal and company-wide financial information. The course is meant to cover an overview of modern techniques aimed at protecting data privacy and security in such applications. Specific Objectives The students will learn the basic cryptographic techniques and their application to obtaining privacy of data in several applications, including cloud computing, statistical databases, distributed computation, and cryptocurrencies. Knowledge and Understanding -) Modern cryptographic techniques and their limitations. -) Techniques for achieving privacy in statistical databases. -) Techniques for designing cryptographic currencies and distributed ledgers. -) Techniques for secure distributed multiparty computation. Applying knowledge and understanding: -) How to select the right cryptographic scheme for a particular application. -) How to design a differentially private mechanism. -) How to program a secure cryptosystem, or a secure smart contract, or a secure cryptographic protocol. Autonomy of Judgment The students will be able to judge the security of the main cryptographic applications. Communication Skills How to describe the security of cryptographic standards, privacy-preserving statistical databases, and blockchains. Next Study Abilities The students interested in research will learn what are the main open challenges in the area, and will obtain the necessary background for a deeper study of the subjects. - VENTURI DANIELE (programma) Il corso offre una panoramica delle tecniche moderne per proteggere la privatezza e la sicurezza dei dati nelle applicazioni digitali. Segue una lista di argomenti. 1) Introduzione alla crittografia: Comunicazione confidenziale, cifratura a chiave segreta e pubblica. Comunicazione autentica, funzioni hash, codici autenticatori di messaggio, firme digitali. 2) Applicazioni: Protocolli per scambio di chiavi ed il protocollo TLS, schemi di autenticazione, password. Sicurezza nel cloud: computazione e memorizzazione. Computazioni verificabili. 3) Privatezza differenziale: Database statistici con privacy. Il meccanismo di Laplace ed il meccanismo esponenziale. Possibilità ed impossibilità. 4) Libri mastri distribuiti e la tecnologia blockchain: Crittovalute ed il protocollo Bitcoin. Smart contracts. 5) Computazione distribuita sicura: Protocolli a due o più giocatori. Conoscenza nulla. Circuiti ingarbugliati di Yao. Protocolli sicuri con maggioranza onesta.   Lucidi del corso preparati dal docente. (Date degli appelli d'esame) 6  INF/01  24  36  -  -  Attività formative caratterizzanti  ENG 1047202 - SIGNAL PROCESSING FOR BIG DATA (obiettivi) This is a methodological course whose objectives are to make the students able to represent information distributed over a graph and to solve complex optimization problems in distributed form. The problems concern the extraction of information from data distributed across various machines and to make inferences about networked data. These goals are achieved by first teaching the theoretical foundations necessary to formulate and solve the above problems and then inviting the students to write Matlab programs implementing the algorithms solving problems of interest. - BARBAROSSA SERGIO (programma) Fondamenti di teoria dei segnali, concetti di base di elaborazione dei segnali e delle immagini, campionamento Teoria algebrica dei grafi, principali proprietà, indici di centralità di un grafo, modelli di grafi aleatori: Erods-Renyi, grafi geometrici aleatori, small-world e grafi scale-free Modelli di indipendenza condizionale, grafi di Markov e Bayes Partizionamento di un grafo, teoria spettrale dei grafi Segnali definiti su grafo, filtraggio e campionamento su grafo, predizione Ottimizzazione convessa, decomposizione primaria e duale, condizioni KKT, compressive sensing ADMM, ottimizzazione distribuita: algoritmi di consenso, algoritmi di condivisione Applicazioni: Algoritmi di machine learning basati su grafi, reti di sensori, allocazione delle risorse radio nelle reti di comunicazione   M. Vetterli et al., “Foundations of Signal Processing” M.E.J. Newman, “Networks: An Introduction” S. Barbarossa, “Signal Processing over Graphs” S. Boyd et al. “Convex Optimization” S. Boyd et al. “Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers” (Date degli appelli d'esame) - DI LORENZO PAOLO (programma) Part I: Signal Processing Definition of signals, signal properties, discrete representations, Fourier transforms, filtering, images, discrete cosine transform, applications to audio signals and images Part II: Processing over graphs Algebraic graph theory, graph properties, connectivity Graph features: degree centrality, eigenvector centrality, PageRank, betweeness, modularity Graph models: random graphs, random geometric graphs, small worlds graphs, scale-free graphs Independence graphs: Markov networks, Bayes networks, Gaussian Markov Random Fields Operations on graphs: partitioning Signals defined on graphs Filtering and sampling signals over graphs Prediction of processes over graphs Part III: Distributed optimization over networks Review of convex optimization Compressive Sensing Algorithms for sparsity constrained problems Primal and dual decomposition Alternating direction method of multipliers Consensus problems Sharing problems Part IV: Examples of application Graph-based methods for machine learning Distributed detection and estimation in wireless sensor networks Radio resource allocation in communication networks Sampling of energy distribution systems   1) Vetterli, Martin, Jelena Kovačević, and Vivek K. Goyal. Foundations of signal processing. Cambridge University Press, 2014. 2) Newman, Mark. Networks: an introduction. Oxford university press, 2010. 3) S. Barbarossa, “Signal Processing over Graphs” 4) S. Boyd and L. Vandenberghe, Convex Optimization, Cambridge University Press, 2004; 5) S. Foucart and R. Holger, A mathematical introduction to compressive sensing, Basel: Birkhäuser, 2013. 6) Boyd, Stephen, et al. "Distributed optimization and statistical learning via the alternating direction method of multipliers." Foundations and Trends® in Machine learning 3.1 (2011): 1-122. 6  ING-INF/03  24  36  -  -  Attività formative caratterizzanti  ENG 10589623 - COMPUTATIONAL DATA ANALYSIS (obiettivi) Le sfide poste dall'analisi hanno evidenziato l'importanza della scalabilità. Algoritmi precedentemente considerati efficienti non sono adeguati al crescere della dimensione e della dimensionalità dei dati oltre certi limiti. Recentemente, tali problemi sono divenuti centrali nella ricerca algoritmica, che ha sviluppato nuovi modelli e paradigmi di progetto. Questo corso verterà sullo studio di alcune di queste tecniche. Gli argomenti di studio includeranno tecniche algoritmiche per importanti problemi di analisi di dati quali il clustering, in scenari in cui i tempi di calcolo debbono essere sub-lineari. Verranno anche considerati scenari on-line e altri in cui soluzioni precedentemente proposte vanno ricalcolate in modo efficiente a seguito di modifiche dinamiche dei dati. Oltre alla presentazione di importanti tecniche algoritmiche, lo scopo del corso è anche introdurre gli studenti alle tecniche di analisi usate per caratterizzarne le prestazioni. - SCHWIEGELSHOHN CHRIS RENE (Date degli appelli d'esame) 6  ING-INF/05  24  36  -  -  Attività formative caratterizzanti  ENG
Gruppo opzionale: GRUPPO OPZIONALE C - (visualizza) 6                1041415 - OPTIMIZATION METHODS FOR MACHINE LEARNING (obiettivi) Knowledge and understanding The aim of the course is to introduce students to the application of optimization techniques to training problems arising in machine learning. Students are expected to gain insight into standard models in Machine Learning (Deep Networks and Support Vector Machines) and into more recent optimization algorithms for determining the parameters (training) of such models that best fit to the available data. Applying knowledge and understanding By the end of the course, students should be able to select the correct model for the problem at hand and either to use standard software specialized to the application and/or to develop their own optimization algorithm. modelli di apprendimento automatico applicati ai casi studio in apprendimento automatico. Making judgements Lectures, practical exercises and project sessions will provide students with the ability to assess the main strengths and weaknesses of the different machine learning models applied to case studies in machine learning. Communication By the end of the course, students are able to point out the main features of a machine learning problem and explain techniques for its solution both with a specialized and a non-specialized audience. These abilities are tested and evaluated in the projects developed in small groups thus encouraging team building and a proactive learning process coupling with collaborative learning. These abilities can also be checked in the final oral exam. Lifelong learning skills Students are expected to develop those learning skills necessary to undertake additional studies on the relevant topics with a high degree of autonomy. During the classes, students are encouraged to work on projects into small groups thus stimulating student activity and engagement. They are pushed to consult supplementary research publications and internet sites to exploit tricks and detailed choices needed to accomplish the tasks effectively. These capabilities are tested and evaluated in the development of the final reports of the projects where students have to discuss the main issues of the addressed problems and their choices to overcome the difficulties, based on the topics and material covered in class. - Erogato presso  1041415 OPTIMIZATION METHODS FOR MACHINE LEARNING in Ingegneria Gestionale LM-31 PALAGI LAURA (Date degli appelli d'esame) 6  MAT/09  24  36  -  -  Attività formative caratterizzanti  ENG
Gruppo opzionale: GRUPPO OPZIONALE D - (visualizza) 12                1047220 - BIOINFORMATICS (obiettivi) This course aims to provide students with a practical and hands-on experience with common bioinformatics tools and databases. Students will be trained in the basic theory and application of programs used for database searching, protein and DNA/RNA sequence analysis, microarray and next generation sequencing data processing and interpretation.It would be expected that after completing this course a student would have a good working knowledge basic bioinformatics tools and databases such as GenBank, BLAST, UCSC genome browser, RNA-seq tools (Tophatt, Cufflinks), and analysis/visualization tools (CIRCOS, Cytoscape). Further, students will understand the basic theory behind these procedures and be able to critically analyze the results of their analysis using such tools. - Tieri Paolo - PETTI MANUELA (programma) - Essentials of: cell biology, genomics and proteomics; the central nervous system and the brain. - Introduction to bioinformatics and computational biology: data, databases, information management and analysis tools. - Introduction to the techniques for studying brain activity and definition of brain connectivity. - Network biology, network medicine, network neuroscience: networks, graphs, biological network analysis and tools. - Seminars on network medicine   Dispense distribuite dai docenti (Date degli appelli d'esame) 6  ING-INF/06  24  36  -  -  Attività formative affini ed integrative  ENG 1056085 - BIG DATA FOR OFFICIAL STATISTICS (obiettivi) What subset of Big Data can be used in the ambit of Official Statistics and what domains of Official Statistics can be enriched through the availability of new data sources. How new data sources can be used in Official Statistics, by taking into account challenges, needs and risks in this exercise. Definition of the role of Big Data in the context of Official Statistics. How to frame the measurement of social, demographic and economic phenomena through Big Data by considering challenges, needs and risks. - MAGGINO FILOMENA (programma) Since the course will be delivered in English, all information will be provided here in that language.   Since the course will be delivered in English, all information will be provided here in that language. (Date degli appelli d'esame) 6  SECS-S/05  24  36  -  -  Attività formative affini ed integrative  ENG 1047216 - DIGITAL EPIDEMIOLOGY (obiettivi) Digital data sources and digital traces of human behaviour have the potential to provide local and timely information about disease and health dynamics at the population level. This course will deal with the opportunities and challenges of mining digital data sources for epidemiological and public health signals and will provide an overview of the state of the art of this emerging field. - PETTI MANUELA (programma) - Essentials of molecular biology - Gene expression data - Dynamic models and identification of gene interactions - Modelling formalisms for a gene network: differential and difference equations, Boolean networks, Bayesian networks, information theory models - Clustering algorithms and their application to gene expression data - Biological networks and precision medicine   Il materiale didattico in formato elettronico (dispense, codice Matlab) sarà distribuito dai docenti. (Date degli appelli d'esame) - FISCON GIULIA (programma) MODULO 2 (3 CFU) La seconda parte del corso si concentrerà sui metodi di teoria delle reti per la bio-medicina. In particolare, il programma prevede una infarinatura di biologia molecolare, descrizione di principali tool basati sulle reti biologiche di co-espressione e interazione proteina-proteina con applicazione a casi di studio reali. Programma dettagliato: - Elementi di Biologia molecolare - Fonti di dati biologiche - Elementi di programmazione R e statistica - Analisi di arricchimento funzionali - Algoritmo basato su reti di correlazione genica - Algoritmo per riposizionamento dei farmaci - Casi di studio   Materiale fornito durante il corso. 6  ING-INF/06  24  36  -  -  Attività formative affini ed integrative  ENG 10589627 - NEURAL NETWORKS FOR DATA SCIENCE APPLICATIONS (obiettivi) Il corso introduce le reti neurali nel contesto delle applicazioni legate al mondo della data science. Dopo un riepilogo dei concetti teorici di base, verranno introdotte alcune delle tecniche ed algoritmi sviluppati negli ultimi anni sotto il nome di “deep learning”, che hanno permesso di applicare con successo le reti neurali a problemi come la computer vision ed il natural language processing. A tal fine saranno introdotte prima le reti convolutive (per l’analisi di immagini), ed in seguito le reti ricorrenti per processare sequenze di dati ed affrontare problemi cosiddetti di ‘sequence to sequence’ (es., traduzione automatica). Ciascun argomento del corso sarà seguito da un laboratorio pratico in cui i concetti saranno sviluppati su casi d’uso concreti tramite l’utilizzo della libreria TensorFlow, rilasciata da Google in open source due anni fa. Risultati di apprendimento attesi: Gli studenti acquisiranno la capacità di sviluppare ed implementare tecniche di apprendimento automatiche (basate sulle reti neurali) per numerosi scenari applicativi, fra cui il riconoscimento di immagini e l’analisi del testo). Inoltre, gli studenti acquisiranno i concetti chiave che gli permetteranno di avvicinarsi a sotto-campi più avanzati in autonomia, come la neural machine translation. A livello pratico, gli studenti impareranno ad utilizzare le più recenti librerie per il deep learning, in particolare con l’utilizzo di TensorFlow (rilasciata da Google). - SCARDAPANE SIMONE (programma) 1. Introduzione e prerequisiti: a. Apprendimento supervisionato e non supervisionato. b. Algoritmi di ottimizzazione stocastica. c. Lab session: Python e NumPy, Jupyter. 2. Deep neural networks a. Architettura di una rete neurale. b. Backpropagation ed allenamento della rete neurale. c. Lab session: introduzione a TensorFlow. 3. Deep learning per la computer vision a. Reti neurali convolutive. b. Scelta della funzione di attivazione. c. Tecniche di regolarizzazione avanzate: dropout e batch normalization. d. Lab session: reti neurali convolutive in TensorFlow. 4. Deep learning per l’analisi del testo a. Word embedding. b. Reti neurali per sequenze (es., reti neurali ricorrenti, LSTM). c. “Sequence to sequence”, traduzione automatica (concetti base). d. Lab session: reti neurali ricorrenti in TensorFlow. 5. Deep reinforcement learning (opzionale) a. Introduzione al reinforcement learning. b. Policy learning con reti neurali (es. deep Q-learning). c. Metodi actor-critic. d. Lab session: OpenAI gym.   1. I. Goodfellow, Y. Bengio and A. Courville, 2017. Deep Learning. MIT Press. 2. Dive into Deep Learning (https://d2l.ai/). (Date degli appelli d'esame) 6  ING-IND/31  24  -  -  -  Attività formative affini ed integrative  ENG
- - A SCELTA DELLO STUDENTE	6		24	36	-	-	Attività formative a scelta dello studente (art.10, comma 5, lettera a)	ITA

Secondo semestre

Insegnamento	CFU	SSD	Ore Lezione	Ore Eserc.	Ore Lab	Ore Studio	Attività	Lingua
Gruppo opzionale: GRUPPO OPZIONALE D - (visualizza) 12                1047218 - EARTH OBSERVATION DATA ANALYSIS (obiettivi) The module aims at providing a general background on the remote sensing systems for Earth Observation from space‐borne platforms and on data processing techniques. It describes, using a system approach, the characteristics of the system to be specified to fulfil the final user requirements in different domains of application. Remote sensing basics and simple wave‐interaction models useful for data interpretation are reviewed together with technical principles of the main remote sensors. The course also provides an overview of the most important applications and bio‐geophysical parameters (of the atmosphere, the ocean and the land) which can be retrieved. The most important techniques for data processing and product generation, also by proposing practical exercises using the computer, are analysed together with an overview of the main Earth Observation satellite missions and the products they provide to the final user. - MARZANO FRANK SILVIO (programma) Ch0. INTRODUCTION. Course presentation. Topics overview. Exam and homeworks. Grouping. [See slides EODA_Ch00_CoursePresentation.pdf] Ch1. STROLLING AROUND EARTH OBSERVATION (Introducing EO between data science and its applications). Data science and its paradoxes: Data scientists, Big data, little data. Earth observation (EO) and data science: Remote sensing and its applications, EO big data and research support services. Data scientist for space sciences: EO opportunities for data scientists, Data scientists skill for EO. Strolling around EO applications: From atmospheric monitoring to climate analysis, From natural hazards to geodesy and geophysics, From urban planning to deforestation surveillance, From environmental to monumental diagnosis. [See slides EODA_Ch01_StrollingEarthObs.pdf] Ch2. EARTH OBSERVATION PRINCIPLES AND CONCEPTS (Overview of EO basic methodologies and techniques). Remote sensing basics: Problem definition and its actors, Target, source, receiver, medium and processes, Inverse problems and retrieval techniques. Electromagnetic radiation basics: Wave fields, electromagnetic spectrum and radiant energy, Wave-matter interaction basic processes and Earth atmosphere, Radiative transfer modeling for Earth observation. Earth observation system basics: EO space segment and ground segments. EO electromagnetic sensors. EO user requirements (radiometric, spectral, spatial, temporal). Remote sensing platforms. Satellite Keplerian orbits (LEO, GEO). [See slides EODA_Ch02_Principles&Concepts.pdf] Ch3. MODELING RADIATION FOR EARTH OBSERVATION (Introducing electromagnetic radiation theory for remote sensing). Wave-matter EM interaction mechanisms: Radiation: intensity, irradiance, exitance and received power, Emission: Planck law, approximations and emissivity, Surface interaction electromagnetic parameters, Volume interaction electromagnetic parameters, Wave reflection and refraction. Radiative transfer theory: Integral-differential equation, Formal integral solution and special cases, Application to absorbing and scattering atmospheres, Application to space and ground remote sensing. Radiation backscatter theory: Wadar equation for single scatterer, Wadar equation for distributed scatterers, Doppler effect and signal statistics. [See slides EODA_Ch03_RadiationModeling.pdf] Ch4. EARTH OBSERVATION SENSORS AND MISSIONS (Introducing EO satellite sensors and missions). Earth observation remote sensors: EO sensor classification and requirements, Passive optical sensors: photocamera principles, Electro-optical sensors: spectroradiometers, interferometers and lidars, Electro-optical sensor scanning systems and geometric distortions, Microwave sensors, imaging radiometers and sounders, Active microwave sensors: altimeters, scatterometers and SARs. Earth observation satellite missions: GEO: EU Meteosat and China Fengyun, LEO: US Aqua and Terra, LEO: US GPM and US/France CALIPSO, LEO: EU MetOP and US Suomi-NPP, LEO: EU Sentinel-1, Sentinel-2 and Sentinel-3, LEO: Italy COSMO-SkyMed and Germany TerraSAR-X, LEO: US DG-High-resolution Worldview. [See slides EODA_Ch04_Sensors&Missions.pdf] Ch5. EARTH OBSERVATION APPLICATIONS (Main applications to Earth science and physically-based techniques). Information content in remote sensing observations: Information content in visible and near-infrared remote sensing, Information content in thermal-infrared and microwave remote sensing. Remote sensing of Earth sea environment: Sea water spectral response, transmittance and reflectance, Visible, near-infrared and thermal-infrared passive remote sensing, Microwave remote sensing: scatterometry, SAR, altimeter and radiometry. Remote sensing of Earth atmosphere: Atmospheric response in the visible-infrared reflective and emissive bands, Profiling radiometric techniques for thermal structure and gas concentration, Water vapor, clouds and precipitation from infrared and microwave radiometers. Remote sensing of Earth solid surface: Vegetation visible-infrared spectral response and retrieval, Rock and surface humidity visible-infrared spectral response and retrieval, Radar and radiometric remote sensing of land surface and emissivity. [See slides EODA_Ch05_EarthObsApplications.pdf] Ch6. EARTH OBSERVATION DATA PROCESSING (Introducing EO data processing and retrieval techniques). EO image data processing: Levels of EO data processing, Color perception and synthesis, Image format and data structure, Image analysis: histogram, contrast, slicing, pseudo-coloring, filtering, Image geocoding: ground control points and resampling. EO inverse problem and retrieval techniques: Inverse and ill-conditioned problems, Regularization, statistical and neural-network solution methods. EO feature extraction and classification: Image feature classification: unsupervised and supervised approach, Feature extraction and principal component analysis, Statistical Bayesian classification method, Thematic map generation process, Image texture exploitation. [See slides EODA_Ch06_DataProcessing.pdf]    Slides of lectures available on the course website (password protected)  Main textobook: Canada Centre, “Fundamentals of remote sensing”, 2008 Freely available on http://www.nrcan.gc.ca/node/9309 Further readings:  Software tools: SNAP official manual and course slides provided by ESA RSS group SNAP is freely downloadable from http://step.esa.int/main/download/ See basic instructions on SNAP available in EODA_ShortGuide-for-SNAP.pdf To actively participate to the ESRIN Lab, download and install SNAP (if needed, you can use a Virtual Machine to exploit the resources available within. ESA Cloud ToolBox facility at http://eogrid.esrin.esa.int/cloudtoolbox after registering). (Date degli appelli d'esame) 6  ING-INF/02  24  36  -  -  Attività formative affini ed integrative  ENG 1056129 - DATA DRIVEN ECONOMICS (obiettivi) A first objective of the course is to provide a basic toolbox for the analysis of agent and group interaction under uncertainty and asymmetric information, and its main consequences on markets enabled from large-scale digital platforms. A second objective of the course will be to provide the basic methods for the use of big data for estimating relevant economic indices. The active participation of students will be stimulated with game-theoretic examples, presentations, simple experiments, case-studies and projects involving the use of real-world data. - Marcucci Juri (Date degli appelli d'esame) - MARZANO RICCARDO 6  SECS-P/02  24  36  -  -  Attività formative affini ed integrative  ENG 1047222 - EFFICIENCY AND PRODUCTIVITY ANALYSIS (obiettivi) This course has the target of providing the students with the modern techniques of measuring quantitatively advanced topics in economic statistics. In particular our focus will be on three main interrelated directions: 1) the analysis of production and efficiency, specifically in the private but also in the public sectors, 2) economic dynamics of sectorial systems founded on micro data, 3) growth, ICT and technology in the modern economy. This course uses statistical methods, both stochastic and deterministic, to analyze topics such as productivity, efficiency and growth at micro, sectorial, and for coherence at macro level. We first take into exam data from firms that will be useful for the mentioned three-levels study, then, as regards the efficiency analysis of productive units, such data will be employed in order to evaluate mergers and acquisitions of plants and firms and management of productive factors. Efficiency will be evaluated from the sides of costs, profits and revenues. As for the sectorial analysis, static and dynamic models will be considered to allow for forecasts and simulations in each sector for variables like production, labour, capital, raw materials, prices and capital gains. As a consequence, an aggregate analysis on the production, growth and prices will follow. We also deal with ICT and technical progress in the production process considering how and if the associated externalities are effective. We will use the following techniques for data analysis: accounting rules for the database, panel data econometrics, time series analysis for systems of equations, methods for differential equation systems. Topics on private and also public sectors will contribute to explain the relationship between economic structure and the actual crisis. Specifically, lectures also include the examinations of cases study concerning the efficiency and productivity analysis on the recent patterns of the banking sector in the international context. - Erogato presso  10589568 EFFICIENCY AND PRODUCTIVITY ANALYSIS in Statistical Methods and Applications LM-82 NESSUNA CANALIZZAZIONE MAGGI BERNARDO 6  SECS-S/03  24  36  -  -  Attività formative affini ed integrative  ENG 10589730 - GEOMATICS AND GEOINFORMATION (obiettivi) The vast majority (a percentage close to 80%) of the currently available data has a geographical connotation, is intrinsically linked to a position; they are therefore named geospatial data. Furthermore, the ever-increasing availability of sensors capable of acquiring geospatial data, allowing the acquisition of larger and larger amounts of data, raises several important issues related to the correct, efficient and effective use of these geospatial big data. This course therefore finds its motivation in the great availability and relevance of geospatial data (in particular big data), and it aims to provide the fundamentals on the main methodologies and techniques currently available for their acquisition, verification, analysis, storage and delivery. Special attention is given to data coming from global navigation satellite systems (GNSS), remote sensing and photogrammetry, volunteered geographic information (VGI) and crowdsourcing, both regarding their analysis and management with freely available software and open source software and their applications. As overall transversal competences, starting from the specific geomatic methods and techniques, students will be trained to develop skills to design both experiments encompassing geospatial big data acquisition, analysis and interpretation, and to use these data to solve interdisciplinary problems also with original approaches. Moreover, students will get the skills enabling the continuous and autonomous update of their methodological and technical competences during their professional life. - CRESPI MATTIA GIOVANNI (programma) Geospatial data and geo big data challenges Geodesy and Cartography: Spatial and temporal reference systems and frames; coordinate systems; practical exercises with online software Positioning with Global Positioning System: positioning and technicalities fundamentals; pseudorange and carrier phase observations modeling and processing; different kinds of GPS surveys, Network Real Time Kinematic surveys; other Global Navigation Satellite Systems; positioning with smartphones; practical exercises with instruments, smartphones and software  Metric information from imagery: photogrammetry fundamentals and models; 3D object reconstruction (surface modeling); digital imagery main features (geometric, temporal, radiometric and spectral resolution); terrestrial, aerial and satellte imagery processing; automated surface modeling and matching techniques; digital terrain models and orthophotos; practical exercises with instruments, smartphones and software Geographical Information Science: data structures, geo big data cloud storage and processing (Google Earth Engine), Volunteered Geographic Information (VGI), crowdsourcing; practical exercises with Google Earth Engine   Slides of the course Reference books and articles: Peter J.G. Teunissen, Oliver Montenbruck (Eds.) (2017). Springer Handbook of Global Navigation Satellite Systems. Springer International Publishing AG. ISBN: 978-3-319-42926-7, DOI 10.1007/978-3-319-42928-1 Karl Kraus (2000). Photogrammetry (vol. 1). Dummler Zhe Jiang, Shashi Shekhar (2017). Spatial Big Data Science - Classification Techniques for Earth Observation Imagery. Springer International Publishing AG. ISBN 978-3-319-60194-6, DOI 10.1007/978-3-319-60195-3 Reference articles: Songnian Li, Suzana Dragicevic, Francesc Antón Castro, Monika Sester, Stephan Winter, Arzu Coltekin, Christopher Pettit, Bin Jiang, James Haworth, Alfred Stein, Tao Cheng (2016). Geospatial big data handling theory and methods: A review and research challenges. ISPRS Journal of Photogrammetry and Remote Sensing 115 (2016) 119–133 Jun Chen, Ian Dowman, Songnian Li, Zhilin Li, Marguerite Madden, Jon Mills, Nicolas Paparoditis, Franz Rottensteiner, Monika Sester, Charles Toth, John Trinder, Christian Heipke (2016). ISPRS Journal of Photogrammetry and Remote Sensing 115 (2016) 3–21 Noel Gorelick, Matt Hancher, Mike Dixon, Simon Ilyushchenko, David Thau, Rebecca Moore (2017). Google Earth Engine: Planetary-scale geospatial analysis for everyone. Remote Sensing of Environment 202 (2017) 18–27 Linda See, Peter Mooney, Giles Foody, Lucy Bastin, Alexis Comber, Jacinto Estima, Steffen Fritz, Norman Kerle, Bin Jiang, Mari Laakso, Hai-Ying Liu, Grega Milˇcinski, Matej Nikšiˇc, Marco Painho, Andrea Podör, Ana-Maria Olteanu-Raimond, Martin Rutzinger (2016). Crowdsourcing, Citizen Science or Volunteered Geographic Information? The Current State of Crowdsourced Geographic Information. ISPRS Int. J. Geo-Inf. 2016, 5, 55; doi:10.3390/ijgi5050055 Maria Antonia Brovelli, Marco Minghini, Giorgio Zamboni (2016). Public participation in GIS via mobile applications. ISPRS Journal of Photogrammetry and Remote Sensing 114 (2016) 306–315 (Date degli appelli d'esame) 6  ICAR/06  24  36  -  -  Attività formative affini ed integrative  ENG
Gruppo opzionale: GRUPPO OPZIONALE B - (visualizza) 18                1056023 - SMART ENVIRONMENTS (obiettivi) Goal of this course is to provide an overview of the large world of wireless and wired technologies that are will be used for the Smart Environments. These technologies will be able to provide infrastructures of networks and digital information used in the urban spaces and smart environments to build advanced applications. Recent advances in areas like pervasive computing, machine learning, wireless and sensor networking enable various smart environment applications in everyday life. The main goal of this course is to present and discuss recent advances in the area of the Internet of Things, in particular on technologies, architectures, algorithms and protocols for smart environments with emphasis on real smart environment applications. The course will present the communication and networking aspects as well as the processing of data to be used for the application design. The course will propose two cases studies in the field of smart environments: Vehicular Traffic monitoring for ITS applications and Network cartography. In both cases instruments, models and methodologies for the design of smart environments applications will be provided. - CUOMO FRANCESCA (programma) The program is divided in four main parts: 1. Enhanced Services by Smart Devices (Introduction to Cyber Physical Systems; Internet of Things (IoT), Manufacturing, Logistic and Supply Chain, Micro-payments by NFC, Intelligent Transport Systems, Indoor Positioning Systems, Pharmaceutical and Healthcare, Building Automation, Smart Home, Smart Cities and Smart GRID) 2. Data Acquisition, Coding, and Aggregation in Smart Environments (Data acquisition: Sampling, quantization; Source coding of multimedia signals (audio, image, video, etc.); Localization technologies and location-aware Services; Georeferenced data collection/aggregation protocols in IoT Networks); 3. Device Communication and Networking (Architectures in the access and in the backbone networks; Basic principles for the wireless networking; Digital modulation. Coding, Static and Dynamic access techniques; Review of TCP/IP networking; LAN/MAN technologies; Ethernet; Wi-Fi; Long Term Evolution and 5G, Low range and low power technologies: RFID, NFC, Bluetooth, Zigbee; LoRAWAN as enabling technologies for IoT; Architectures and protocols for virtual and augmented reality environments thought Voice and Video over IP, Dynamic Adaptive HTTP Streaming, Video 360°); 4. Practical examples of Data Processing for Smart Environments (Vehicular Traffic monitoring: Data-driven vs. Model-driven approaches; Online vehicular traffic prediction and anomaly detection; Analysis derived from real data and emulation platforms; Examples to be defined during the course projects)   Lucidi del corso e articoli disponibili sul sito web https://sites.google.com/a/uniroma1.it/francescacuomo/didattica (Date degli appelli d'esame) - BIAGI MAURO 6  ING-INF/03  24  36  -  -  Attività formative caratterizzanti  ENG 10589621 - ADVANCED MACHINE LEARNING - GALASSO FABIO (programma) The course would present advanced concepts of machine learning and their application in computer vision via deep neural network (DNN) models. It would include theory and practical coding, as well as a final hands-on project. In a first part of the course, I would introduce state-of-the-art DNN models for classification, showing how to estimate which objects are within an image. I would then showcase regression, as applied to detection (where the objects are in the image), pose estimation (whether people stand, sit or crunch) and re-identification (estimating a unique vector representation for each person). I would further discuss DNNs for multi-task objectives (joint detection, pose estimation, re-identification, segmentation, depth estimation etc). This first part would include DNNs which apply to video sequences, by leveraging memory (e.g. LSTMs) or attention (Transformers). In a second part of the course, I would discuss generalization and the effective use of labelled and unlabelled data for learning. Further to transfer learning (how pre-trained models may be deployed for other tasks), I would discuss multi-modal (with different sensor modalities such as depth or thermal cameras) and self-supervision (e.g. training the DNN model by solving jigsaw puzzles) to auto-annotate large amounts of data. Finally, I would present domain adaptation (e.g. apply daytime-detectors for night vision) and meta-learning, a most recent framework to learn how to learn a task, e.g. online or from little available data. Course website: https://sites.google.com/di.uniroma1.it/aml-19-20 Course Google group: https://groups.google.com/a/di.uniroma1.it/forum/#!forum/aml_19_20   Slides and coding scripts would be distributed after lectures, as well as references to online material including papers and blogs. Reference books: - for Machine Learning: - Christopher Bishop, 2006. Pattern Recognition and Machine Learning. - for Deep learning: - Ian Goofellow, Yoshua Bengio, Aaron Courville, 2017. Deep Learning. - Andrew Ng, 2019. Machine Learning Yearning - for Computer Vision: - Richard Szeliski, 2010. Computer Vision: Algorithms and Applications Reference book for Python: - Allen B. Downey, 2015. Think Python: How to Think Like a Computer Scientist Online tutorials for Python: https://docs.python.org/3/tutorial/ Online tutorials for Pytorch: https://pytorch.org/tutorials/ (Date degli appelli d'esame) 6  INF/01  24  36  -  -  Attività formative caratterizzanti  ENG
Gruppo opzionale: GRUPPO OPZIONALE C - (visualizza) 6                1056087 - STATISTICS FOR STOCHASTIC PROCESSES (obiettivi) Lo scopo principale del corso è quello di insegnare le più popolari metodologie di inferenza non parametrica, soprattutto da un punto di vista bayesiano. Lo studente, alla fine del corso, sarà in grado di utilizzare processi di Dirichlet e loro ramificazioni nei più consueti contesti inferenziali di data science - LISEO BRUNERO (programma) Topics: A1:Scambiabilità e teorema di de Finetti A2: Processi di Dirichlet: Definizione, proprietà, applicazioni); Altri processi a priori. A3: Misture di processi di Dirichlet – metodologia, formulazione del modello, rappresentazioni aletrnative, simulazione a posteriori. A4: Applicazioni in modelli specifici A5: Processi a priori con dipendenza (se il tempo lo consente A6: processi Gaussiani   Slides e appunti a cura del docente Fundamentals of Nonparametric Bayesian Inference (Cambridge Series in Statistical and Probabilistic Mathematics) 1st Edition by Subhashis Ghosal, Aad van der Vaart (2018) Bayesian Nonparametric Data Analysis (Springer) 2015 by Peter Müller, Fernando Andres Quintana, Alejandro Jara, Tim Hanson (Date degli appelli d'esame) 6  SECS-S/01  24  36  -  -  Attività formative caratterizzanti  ENG
AAF1149 - altre conoscenze utili per l'inserimento nel mondo del lavoro (obiettivi) The specific aim is to enable the student to assist him with the more specific knowledge for inclusion in the future world of work. - TARDELLA LUCA (Date degli appelli d'esame)	3		-	-	-	-	Ulteriori attività formative (art.10, comma 5, lettera d)	ITA
AAF1022 - PROVA FINALE (obiettivi) The student will present and discuss the results of a technical activity, producing a written thesis supervised by a professor and showing the ability to master the methodologies of data science	24		-	-	-	-	Per la prova finale e la lingua straniera (art.10, comma 5, lettera c)	ITA

• Contatti
• Accessibilità
• Mappa del sito

© Università degli Studi di Roma "La Sapienza" - Piazzale Aldo Moro 5, 00185 Roma T (+39) 06 49911 CF 80209930587 PI 02133771002

Obiettivi formativi