2. Professors
  3. CARDONE Angelamaria
  4. Teaching

NUMERICAL METHODS FOR DATA ANALYSIS IN CYBERSECURITY

Angelamaria CARDONE NUMERICAL METHODS FOR DATA ANALYSIS IN CYBERSECURITY

0522700012
COMPUTER SCIENCE
EQF7
CYBERSECURITY AND CLOUD TECHNOLOGIES
2024/2025

YEAR OF COURSE 2
YEAR OF DIDACTIC SYSTEM 2023
AUTUMN SEMESTER
CFUHOURSACTIVITY
432LESSONS
216LAB
ANGELAMARIA CARDONE T Curriculum
DAJANA CONTE Curriculum
Objectives
THE TEACHING AIMS TO PROVIDE THE BASIC THEORETICAL KNOWLEDGE AND PROMOTE THE DEVELOPMENT OF SKILLS IN THE USE OF NUMERICAL METHODS FOR DATA ANALYSIS IN THE CYBERSECURITY FIELD.

KNOWLEDGE AND UNDERSTANDING
THE STUDENT WILL ACQUIRE KNOWLEDGE AND UNDERSTANDING REGARDING:
• THE STUDY OF INFORMATION: RESEARCH AND ANALYSIS OF TEXTS, SELECTION OF RELEVANT INFORMATION, WAYS OF DISSEMINATING INFORMATION AND DISINFORMATION;
• NUMERICAL LINEAR ALGEBRA METHODS AIMED AT ACQUIRING RELEVANT INFORMATION AND STUDYING COMPLEX NETWORKS;
• THE MAIN CENTRALITY METRICS IN A GRAPH, AIMED AT UNDERSTANDING THE NODES OF A NETWORK CAPABLE OF ACQUIRING OR SUPPLYING MORE INFORMATION;
• SOME DIFFERENTIAL MODELS FOR THE DISSEMINATION OF INFORMATION AND MALWARE;
• THE MAIN MATHEMATICAL FOUNDATIONS OF MACHINE LEARNING, WITH PARTICULAR ATTENTION TO ARTIFICIAL NEURAL NETWORKS AND THE BASIC NUMERICAL TECHNIQUES FOR THEIR APPLICATION.

ABILITY TO APPLY KNOWLEDGE AND UNDERSTANDING
THE STUDENT WILL BE ABLE TO
• ANALYZE NUMERICAL METHODS;
• DEVELOP AND USE MATHEMATICAL SOFTWARE FOR THE SELECTION OF RELEVANT INFORMATION, FOR THE CALCULATION OF DIFFERENT CENTRALITY METRICS IN A GRAPH, FOR MODELING THE DIFFUSION OF INFORMATION

MAKING JUDGMENTS
THE STUDENT WILL BE ABLE TO:
• SELECT THE MOST SUITABLE NUMERICAL METHOD FOR THE PROBLEM UNDER EXAMINATION BY ANALYZING THE CHARACTERISTICS OF THE PROBLEM ITSELF;
• ANALYZE VARIOUS CASE STUDIES, RELATING TO TOPICS COVERED DURING THE COURSE, USING APPROPRIATE MATHEMATICAL MODELS, COMPARING THE RESULTS OBTAINED FROM THESE MODELS WITH THE AVAILABLE DATA.

COMMUNICATION SKILLS
THE STUDENT WILL BE ABLE TO:
• CLEARLY PRESENT THE THEORETICAL CONTENTS RELATING TO THE TOPICS COVERED DURING THE FRONTAL LESSONS;
• CARRY OUT PROJECTS IN GROUPS, TO SOLVE PROBLEMS RELATED TO THE TOPICS OF THE COURSE, AND REPORT ON THE ACTIVITY CARRIED OUT.

LEARNING SKILLS
THE STUDENT WILL BE ABLE TO:
• USE TRADITIONAL BIBLIOGRAPHIC TOOLS AND IT RESOURCES FOR ANALYSIS AND ARCHIVING.
Prerequisites
ELEMENTS OF DISCRETE MATHEMATICS AND MATRICIAL CALCULUS.
Contents
1. FUNDAMENTALS OF MATRICIAL CALCULUS. DATA ANALYSIS THROUGH SVD MATRIX FACTORIZATION: PRINCIPAL COMPONENT ANALYSIS, LATENT SEMANTIC ANALYSIS. IMAGE COMPRESSION. RECOMMENDATION SYSTEMS. GRADIENT AND NEAREST NEIGHBORS TECHNIQUES. (12 HOURS)
2. MATRICIAL METHODS FOR THE ANALYSIS OF COMPLEX NETWORKS: GRAPHS AND MATRIXES, CONNECTIVITY AND ADJACENCE NETWORKS. MEASURES OF CENTRALITY AND COMMUNICABILITY, SIMILARITY AND DISTANCE BASED ON SPECTRAL TECHNIQUES AND MATRIX FUNCTIONS. HITS AND PAGE RANK METHODS FOR RANKING OF COMPLEX NETWORKS. (12 HOURS)
3. DISINFORMATION AND FAKE NEWS. EPIDEMIOLOGICAL MODELS FOR THE ANALYSIS OF INFORMATION DIFFUSION. APPLICATION OF EPIDEMIOLOGICAL MODELS ALSO FOR THE SPREAD OF MALWARE. (12 HOURS)
4. NUMERICAL TECHNIQUES FOR DEEP LEARNING: ACTIVATION FUNCTIONS AND COST FUNCTIONS, GRADIENT DESCENT, STOCHASTIC GRADIENT DESCENT. LINEAR MODELS: LEAST SQUARES. ARTIFICIAL NEURAL NETWORKS, BACKPROPAGATION, PHYSICS-INFORMED NEURAL NETWORKS. (12 HOURS)
Teaching Methods
6 CFU, 48 HOURS FOR:
- LECTURES (32 HOURS)
- LABORATORY (16 HOURS)

THE CLASSROOM LESSONS WILL PRESENT THE THEORETICAL FOUNDATIONS WHICH WILL ALLOW STUDENTS TO APPLY THIS KNOWLEDGE IN THE LABORATORY LESSONS.
FOR EACH OF THE TOPICS COVERED, SITUATIONS OF INTEREST IN REALITY WILL BE PRESENTED WHICH REQUIRE THE USE OF THE NUMERICAL METHODOLOGIES STUDIED.
Verification of learning
- PRACTICAL TEST: EXECUTION OF THE MATHEMATICAL SOFTWARE DEVELOPED OR USED IN THE LABORATORY LESSONS. THIS TEST AIMS TO VERIFY THE STUDENT'S ABILITY TO SOLVE SIMPLE PROBLEMS RELATED TO DATA ANALYSIS AND TO COMPARE THE PERFORMANCE OF DIFFERENT CODES.
- ORAL INTERVIEW ON THEORETICAL CONTENTS TO VERIFY KNOWLEDGE OF THE BASIC NOTIONS OF THE NUMERICAL METHODS COVERED FOR PROBLEMS RELATING TO DATA ANALYSIS.

- FOR STUDENTS WHO ATTEND THE COURSE, TWO ONGOING EXEMPTION TESTS ARE PROVIDED, ACCORDING TO THE SAME METHOD AS THE EXAM.
Texts
A. MAIN REFERENCE TEXTS
1.BRUNTON, S., & KUTZ, J. (2019). DATA-DRIVEN SCIENCE AND ENGINEERING: MACHINE LEARNING, DYNAMICAL SYSTEMS, AND CONTROL. CAMBRIDGE UNIVERSITY PRESS.
2.DAVID C. ANASTASIU, ET AL., BIG DATA AND RECOMMENDER SYSTEMS, 2016.
3.CHRISTOPHER R. ABERGER, RECOMMENDER: AN ANALYSIS OF COLLABORATIVE FILTERING TECHNIQUES, 2016,
4.N. LANGVILLE, C. D. MEYER: GOOGLE'S PAGERANK AND BEYOND. PRINCETON UNIV. PRESS, 2006.
5.CATHERINE F. HIGHAM AND DESMOND J. HIGHAM, DEEP LEARNING: AN INTRODUCTION FOR APPLIED MATHEMATICIANS | SIAM REVIEW
6.MICHELE BENZI, PAOLA BOITO, MATRIX FUNCTIONS IN NETWORK ANALYSIS - BENZI - 2020 - GAMM-MITTEILUNGEN - WILEY ONLINE LIBRARY
7.MARTIN ATZMUELLER, RUSHED KANAWATI, EXPLAINABILITY IN CYBER SECURITY USING COMPLEX NETWORK ANALYSIS: A BRIEF METHODOLOGICAL OVERVIEW (ACM.ORG)
8.MICHAEL MUHLMEYER, SHAURYA AGARWAL, INFORMATION SPREAD IN A SOCIAL MEDIA AGE, MODELING AND CONTROL, WITH MATLAB EXAMPLES AND CASE STUDIES, TAYLOR & FRANCIS, 2021.
9.NETWORKS: AN INTRODUCTION, MARK E. J. NEWMAN. 2010, OXFORD UNIVERSITY PRESS
10. SLIDES OF THE COURSE HTTP://ELEARNING.INFORMATICA.UNISA.IT


B. SECONDARY
-V. COMINCIOLI, METODI NUMERICI E STATISTICI PER LE SCIENZE APPLICATE, MILANO, AMBROSIANA, 1992.
-T. HASTIE, R. TIBSHIRANI J. FRIEDMAN: THE ELEMENTS OF STATISTICAL LEARNING: DATA MINING, INFERENCE, AND PREDICTION. SECOND EDITION, 2009
-I.T. JOLLIFFE, PRINCIPAL COMPONENT ANALYSIS, SECOND EDITION, SPRINGER, 2002
-A FIRST COURSE IN NETWORK THEORY. ESTRADA, ERNESTO, AND PHILIP A. KNIGHT. OXFORD UNIVERSITY PRESS, USA, 2015.
More Information
ANCARDONE@UNISA.IT; DAJCONTE@UNISA.IT
Lessons Timetable

  BETA VERSION Data source ESSE3 [Ultima Sincronizzazione: 2024-11-29]