SOFTWARE ENGINEERING FOR ARTIFICIAL INTELLIGENCE

International Teaching SOFTWARE ENGINEERING FOR ARTIFICIAL INTELLIGENCE

0522500139
COMPUTER SCIENCE
EQF7
COMPUTER SCIENCE
2021/2022

YEAR OF COURSE 1
YEAR OF DIDACTIC SYSTEM 2016
SECONDO SEMESTRE
CFUHOURSACTIVITY
630LESSONS
Objectives
KNOWLEDGE AND UNDERSTANDING. STUDENTS WILL HAVE KNOWLEDGE OF THE METHODOLOGIES AND TECHNIQUES FOR THE ANALYSIS, DESIGN, AND VERIFICATION OF ARTIFICIAL INTELLIGENCE-BASED SOFTWARE SYSTEMS AND, MORE PARTICULARLY, OF SYSTEMS THAT REQUIRE THE INTENSIVE USAGE OF MACHINE LEARNING APPROACHES IN THE CONTEXT OF COMPLEX AND CRITICAL SOFTWARE SYSTEMS.

APPLYING KNOWLEDGE AND UNDERSTANDING. STUDENTS WILL BE ABLE TO ENGINEER AND OPERATE ON ARTIFICIAL INTELLIGENCE PIPELINES, BEING ABLE TO ACT OVER AN ENTIRE LIFE CYCLE BASED ON MLOPS, I.E., FROM THE DEFINITION TO THE MONITORING OF THOSE PIPELINES.
Prerequisites
THE STUDENTS MUST HAVE PREVIOUS KNOWLEDGE OF SOFTWARE ENGINEERING CONCEPTS – SOFTWARE EVOLUTION IN PARTICULAR – AND ARTIFICIAL INTELLIGENCE – IN THIS CASE, IT IS ENOUGH TO HAVE ACQUIRED BASIC SKILLS.
Contents
THE CONTENT OF THE COURSE IS ORGANIZED AS FOLLOW:

- DATA EXTRACTION, VALIDATION, AND PRE-PROCESSING, WITH A PARTICULAR EMPHASIS ON BIG DATA EXTRACTION.
- ARTIFICIAL INTELLIGENCE MODELS TRAINING AND CONFIGURATION;
- FUNDAMENTALS OF SOFTWARE ANALYTICS FOR ARTIFICIAL INTELLIGENCE MODELS, WITH A PARTICULAR EMPHASIS ON VALIDATION/VERIFICATION, SECURITY/PRIVACY, LEARNING BIAS, AND MODEL EXPLAINABILITY;
- ARTIFICIAL INTELLIGENCE MODELS VERSIONING;
- ARTIFICIAL INTELLIGENCE MODELS DEPLOYMENT AND MONITORING;
- ARTIFICIAL INTELLIGENCE MODELS MANAGEMENT AND DOCUMENTATION.
Teaching Methods
THE COURSE INCLUDES 30 HOURS OF FRONTAL LECTURES (6 ECTS) TO TRANSFER THE KNOWLEDGE RELATED TO THE THEORETICAL/METHODOLOGICAL CONTENTS AND TO THE TOOLS REQUIRED TO CARRY OUT A PROJECT.

THE STUDENTS WILL CARRY OUT A PROJECT (INDIVIDUALLY OR IN A GROUP OF MAXIMUM 4 STUDENTS) TO TRAIN ON THE PRACTICAL ACTIVITIES OF THE COURSE. THE EFFORT REQUIRED FOR THE PROJECT ACTIVITIES IS ABOUT 30 HOURS.
Verification of learning
LEARNING ASSESSMENT IS BASED ON AN EXAM WITH GRADES ON A SCALE OF 30. THE EXAM CONSISTS OF (1) AN INDIVIDUAL OR GROUP PROJECT AND (2) AN ORAL EXAMINATION.

DELIVERING THE PROJECT DOCUMENTATION IS PREPARATORY FOR THE ORAL EXAMINATION.

THE ORAL EXAMINATION IS BASED ON QUESTIONS AND DISCUSSION ABOUT THE ISSUES RAISED DURING THE DEVELOPMENT OF THE PROJECT AND ON THE THEORETICAL AND METHODOLOGICAL TOPICS OF THE COURSE. IT AIMS AT VERIFYING THE LEVEL OF KNOWLEDGE ACQUIRED BY THE STUDENT ON THE THEORETICAL AND METHODOLOGICAL TOPICS OF THE COURSE, HOW THE METHODS PRESENTED DURING THE COURSE HAVE BEEN APPLIED WITHIN THE PROJECT, THE ACTUAL CONTRIBUTION GIVEN TO THE PROJECT, AND THE LEVEL OF CORRECTNESS AND COMPLETENESS OF THE PROJECT DOCUMENTATION. THE ORAL EXAMINATION ALSO AIMS AT VERIFYING THE CAPABILITY OF AUTONOMOUSLY ORGANIZING THE PRESENTATION BY USING THE CORRECT TERMINOLOGY AND THE CAPABILITY OF PROPERLY MOTIVATING AND DISCUSSING THE PROJECT CHOICES.
Texts
- H. HAPKE, C. NELSON, “BUILDING MACHINE LEARNING PIPELINES”, O’REALLY.
- A. BURKOV, “MACHINE LEARNING ENGINEERING”, PAPERBACK.
More Information
ATTENDING THE COURSE IS NOT MANDATORY BUT STRONGLY RECOMMENDED. STUDENTS MUST BE READY TO ATTEND THE COURSE ACTIVELY, THROUGH THE INTERACTION WITH THE LECTURER AS WELL AS THE INDIVIDUAL STUDY OF THE MATERIAL TAUGHT DURING THE LECTURES. A SATISFACTORY PREPARATION WHICH LEADS TO PASSING THE EXAM WILL CONSIST OF AN AVERAGE INDIVIDUAL STUDY OF TWO HOURS FOR EACH HOUR OF LECTURE AND AN AVERAGE OF ONE HOUR DEVOTED TO THE DEVELOPMENT OF THE PROJECT.

BY DESIGN, THE COURSE EXPECTS A STRONG PREDISPOSITION TO LEARNING NEW SOFTWARE INSTRUMENTS FOR THE DEVELOPMENT OF ARTIFICIAL INTELLIGENCE PIPELINES (E.G., JUPYTER NOTEBOOK, KUBERFLOW, ETC.). THE DIDACTIC MATERIAL WILL BE MADE AVAILABLE ON THE E-LEARNING PLATFORM OF THE DEPARTMENT.

LECTURER:

FABIO PALOMBA
E-MAIL: FPALOMBA@UNISA.IT
SITO WEB: HTTPS://FPALOMBA.GITHUB.IO
  BETA VERSION Data source ESSE3