## ESTIMATION AND DATA ANALYSIS WITH APPLICATIONS

Teaching in italian
ESTIMATION AND DATA ANALYSIS WITH APPLICATIONS
Teaching
ESTIMATION AND DATA ANALYSIS WITH APPLICATIONS
Subject area
ING-INF/04
Reference degree course
COMPUTER ENGINEERING
Course type
Laurea Magistrale
Credits
9.0
Teaching hours
Ore Attività frontale: 81.0
Academic year
2018/2019
Year taught
2019/2020
Course year
2
Language
INGLESE
Curriculum
PERCORSO COMUNE
Reference professors for teaching
Location
Lecce

### Teaching description

Sufficiency in calculus, probability theory, linear algebra.

This course offers a broad overview of fundamental and emerging topics in the area of estimation theory and data analysis; furthermore, a set of applications are illustrated in the fields of robotics, multi-agent and cyber-physical systems, social systems and electric networks. It is aimed at providing principles and tools to state and solve estimation problems in technological systems, and the solution is numerically sought with the aid of a suitable software (Mathworks Matlab is used in the course).

Learning Outcomes; after the course the student should be able to:

(Conoscenze e comprensione) Describe and explain the main peculiarities (both advantages and disadvantages) of each mathematical framework for the estimation problems considered in the course.

(Capacità di applicare conoscenze e comprensione)+ (Abilità comunicative) + (Autonomia di giudizio) Be aware of, describe and explain practical problems of bad data gathering and robustness issues in the framework of estimation theory.

(Capacità di applicare conoscenze e comprensione)+ (Capacità di apprendimento) For a given practical problem at hand, be able to state an estimation problem in a natural mathematical setting, either stochastic or deterministic, based on the problem assumptions.

(Capacità di applicare conoscenze e comprensione) +(Abilità comunicative) + (Autonomia di giudizio) Build a simulation framework to find a computer-aided solution of the stated mathematical problem with the use of a suitable software.

(Abilità comunicative)+(Capacità di apprendimento) Willing students may hold a seminar on an application of interest where to apply the methodologies developed along the course.

Lezioni frontali svolte in aula dal docente tramite l'ausilio di gesso e lavagna. Nel corso delle lezioni saranno occasionalmente illustrati e discussi  software commerciali.

The exam is an oral discussion (including possibly one written exercise) and it is aimed to determine to what extent the student has: 1) the ability to identify and use data to formulate responses to well-defined problems, 2) problem solving abilities to seek a solution through an algorithm. Additionally, willing students may have a seminar on an application of interest where the methodologies of the course are applied.

Introduction. Mathematical background and connections with other courses (2 hours). Stochastic Estimators: definitions, properties, performances and fundamental limitations. Foundations of maximum likelihood estimation (10 hours). The Bayesian approach to the estimation problem (7 hours). Kalman filter: discrete-time stochastic state models, (two-steps) structure, computation of the optimal gain, the alternative geometric approach. Steady–state behavior. Extended Kalman Filter (16 hours). Applications of Kalman Filter (6 hours). Set membership estimation: introduction, fundamental results and theorems (8 hours). Set membership estimation: some applications (4 hours). Robust estimation: introduction, fundamental definitions, estimator classes and performances (7 hours). Data driven by unknown external entities: vulnerability analysis, resilient estimator design (6 hours). Applications of the previous issues and results to various fields (3 hours). Data analysis: mathematical tools, foundations. Elements of clustering and classification (7 hours). The electric power system state estimation. Overview of Electric Power System State Estimation techniques. (5 hours).

 Ljung, Lennart. "System Identiﬁcation: Theory for the user" Englewood Cliffs, 1987.

 Anderson, Brian DO, and John B. Moore. "Optimal Filtering" (1979).

 Milanese, M., Norton, J., Piet-Lahanier, H., & Walter, É. (Eds.). (2013). Bounding approaches to system identification. Springer Science & Business Media.

 Zaki, Mohammed J., and Wagner Meira Jr. “Data mining and analysis: fundamental concepts and algorithms”, Cambridge University Press, 2014.

Semester
Secondo Semestre (dal 02/03/2020 al 05/06/2020)

Exam type
Non obbligatorio

Type of assessment
Orale - Voto Finale

Course timetable
https://easyroom.unisalento.it/Orario