Program Overview

COGNITIVE TELECOMMUNICATION SYSTEMS

The COGNITIVE TELECOMMUNICATION SYSTEMS program is designed to provide Master's students with basic and advanced concepts on the design of methods and techniques for data-driven self-awareness in autonomous artificial agents. The program focuses on signal processing, data fusion, and machine learning under a Bayesian perspective.

OVERVIEW

This course aims to introduce students to the design of cognitive telecommunication systems, emphasizing the development of self-awareness in autonomous agents. The program covers laboratory applications and agent design, integrating theoretical activities with practical experiences.

AIMS AND CONTENT

The module aims to provide theory and techniques for architectural and functional design of interactive cognitive dynamic systems. Topics include data fusion, multilevel Bayesian state estimation, and their application to cognitive video and radio domains.

LEARNING OUTCOMES

• Basic and advanced knowledge on design of telecommunication systems frameworks for context-aware multisensorial processing of signals and data in cognitive agents
• Knowledge on methods and techniques for acquisition, joint representation, and processing of proprioreceptive and exteroceptive multisensorial signals in cognitive dynamic agents
• Knowledge on methods and techniques for Multisensor Data Fusion: coupled hierarchical processing of multisensorial signals
• Knowledge on Machine Learning methods and techniques based on Cognitive Dynamic Systems theory for Situation awareness and Self-awareness in artificial cognitive agents

PREREQUISITES

• Probability theory
• Random Processes
• Signal theory

TEACHING METHODS

The course is divided into two parts: lectures and laboratory experiences. Lectures will cover theoretical concepts, while laboratory experiences will involve the application of programs in Matlab framework.

SYLLABUS/CONTENT

The syllabus includes:

• Introduction to cognitive dynamic systems
• Bayesian Networks
• Bayesian inference: filtering, smoothing, prediction, and update
• Hierarchical representations and Switching models
• Generative models: representation and inference
• Machine learning for HDBNs
• Incremental learning of switching generative models
• Application and integration of course concepts by multisensory dataset processing

RECOMMENDED READING/BIBLIOGRAPHY

Slides of all lectures will be made available, along with recommended books and research papers to integrate concepts described during frontal activity.

TEACHERS AND EXAM BOARD

The course is taught by Carlo Regazzoni, with Pamela Zontone as part of the exam board.

EXAM DESCRIPTION

The exam consists of a written and an oral part. The written part requires the presentation of a report or poster describing a set of activities and results, demonstrating knowledge and capabilities acquired during the course.

ASSESSMENT METHODS

The exam aims to assess the level of knowledge acquired, practical and integration capabilities, and the ability to motivate performed choices and obtained results.

EXAM SCHEDULE

The exam schedule is available, with multiple dates throughout the year.

LESSONS

The timetable for this course is available, and lessons start on a specified date.

FURTHER INFORMATION

Dataset will be assigned at least two weeks before the exam, and reports must be presented on the Monday before the exam date. Oral admission will be communicated a day before the oral exam.