Coordinators: Prof. Bruno Ciciani and Prof. Christian Napoli

Course Objectives

The aim of the course is to offer the students a series of seminars, held by a group of teachers and/or experts in the single topics, on the latest trends in research and in the practice of engineering in computer science,

This year the Seminars in Advanced topics in engineering in computer science course focuses on two main topics:

• high-end computing architectures and programming paradigms for high performance computing

• machhine learning and features representation techniques for data science.
  

The course is organized in 24 seminars, 12 are related to the first topic and the other 12 to the second ones.

The first topic focus on modern dependable and highly performant computing architectures, from single-chip level to large computer clusters, with special attention on multicore systems and massively-parallel architectures. Programming approaches for exploiting the computing power of these architectures are presented and discussed, with special attention on concurrent and distributed programming.

The second topic focus on classical and modern machine learning approaches for data science. Several models of artificial neural networks will be presented and discussed along with the related advantages and drawbacks. From the standard feedforward model the lectures will move to recurrent and dynamical networks, as well as probabilistic network and deep learning models for classification. Several algorithms for supervised and unsupervised training will be devised, focusing on the training models that are suitable for the most common neural architectures. An important part of the lectures will be reserved to describe the most effective techniques of feature extraction starting from signal analysis to entropy evaluation models for information compression and noise reduction, as well as for data filtering and feature representation.

Exam

A presence sheet will be available in each seminar to register presence.

There are three exam modalities:

- Attendance at least of the 75% (18 over 24) of all the seminars:

* PAPER PRESENTATION REGARDING ONE OF THE TOPIC PRESENTED (free choice).

or

* DO A PROJECT REGARDING ONE OF THE TOPICS (free choice, but contact a teacher).

- Attendance at least of the 75% (9 over 12) of the seminars related to one of the two topics:

* PAPER PRESENTATION (the paper is assigned by the Teacher).

or

* DO A PROJECT REGARDING ONE OF THE TOPICS (free choice, but contact a teacher).

- Other cases:

* PAPER PRESENTATION FOR BOTH TOPICS (assigned by the teachers).

or

* DO A PROJECT FOR BOTH TOPICS (free choice, but contact the teachers).

Temporary (tentative) calendar: