B8376 - METODI STATISTICI PER L'INGEGNERIA DELL'INFORMAZIONE LM

Learning outcomes

At the end of the course, the student will have advanced competencies in detection, estimation, and modeling of random processes, with a focus on applications in the field of information engineering. The student will have the necessary knowledge to design advanced systems for parameter detection and estimation, leveraging statistical techniques for signal and data analysis. Particular attention is given to real-world case studies, such as those related to radar systems and telecommunications.

Course contents

Probability and random processes refresher

Statistical estimation (ML, MAP, MMSE, Fisher Information Matrix, Cramer-Rao Bound)

Decision theory (Bayesian and Neyman-Pearson tests, ROC curves)

Statistical filtering (Kalman filter, EKF, particle filter)

Statistical machine learning foundations

Applications to radar signal processing

Wireless and MIMO channel estimation

Hands-on projects using Python/Matlab

Readings/Bibliography

- Lecture notes

- S. Kay, Fundamentals of Statistical Signal Processing, Volume I: Estimation Theory, and Volume II: Detection Theory. Prentice Hall 1998 - Van Trees, Detection, Estimation, and Modulation Theory, Part I, second edition. Wiley 2010.

- T. Hastie, R. Tibshirani, and J. Friedman, The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd ed. Springer, 2017. Free online version

Teaching methods

The course includes:

• Lectures for the presentation of theoretical content, supported by slides and blackboard explanations.

• Guided exercises aimed at the practical application of statistical methods to real-world problems.

• Laboratory sessions, where small projects and simulations will be developed using programming languages such as Python or Matlab, focusing on estimation, filtering, and classification.

Active student participation is encouraged through questions, discussions, and short in-class exercises.

Assessment methods

Assessment is carried out exclusively through a written exam, typically lasting 2 hours, and includes:

• Theoretical questions (open or short answer) on the fundamental concepts covered in the course.

• Practical exercises involving calculation, modeling, and simulation of estimation, filtering, and decision-making problems.

No oral exam is foreseen.
The final grade is expressed on a scale of 30, with honors (cum laude) possible.

Teaching tools

• Lecture slides and notes, available on the university’s e-learning platform.

• Reference handouts and selected scientific articles for in-depth study.

• Python/Matlab notebooks provided by the instructor for exercises and simulations.

• Forum or Teams channel for ongoing support, answering questions, and interaction between students and the instructor.

Office hours

See the website of Marco Chiani