Instructor: Prof. Marco Levorato, University of California Irvine (UCI), US Bio: Marco Levorato joined the Computer Science department at UCIrvine in August 2013. Between 2010 and 2012, He was a post-doctoral researcher with a joint affiliation at Stanford and the University of Southern California working with prof. Andrea Goldsmith and prof. Urbashi Mitra. From January to August 2013, he was an Access post-doctoral affiliate at the Access center, Royal Institute of Technology, Stockholm. He is a member of the ACM, IEEE and IEEE Comsoc society. His research interests are focused on next-generation wireless networks, signal processing, cyber-physical systems, smart city and smart energy systems. He has co-authored over 75 technical articles on these topics, including the paper that has received the best paper award at IEEE GLOBECOM (2012). He completed the PhD in Electrical Engineering at the University of Padova, Italy, in 2009. He obtained the B.S. and M.S. in Electrical Engineering summa cum laude at the University of Ferrara, Italy in 2005 and 2003, respectively. In 2016, he received the UC Hellman Foundation Award for his research on Smart City IoT infrastructures.

-

This course consists of blocks of lectures and projects on the topics of machine learning concepts and algorithms, and their application to the field of autonomous mobile systems and the associate computed vision tasks. In its first part, the course will cover basic machine learning concepts and algorithms including fundamental deep neural network architectures (multi layer perceptron, convolutional neural networks and recurrent neural networks). Then, as a use case, we will describe the perception pipeline for autonomous vehicles and the associated computer vision tasks and neural models. Then, we will discuss the challenges of supporting the execution of complex neural models on resource constrained mobile devices, and current techniques to address this issue in the context of 5G systems. The sections of the course on machine learning are based on the book Deep Learning by Ian Goodfellow, Yoshua Bengio, Aaron Courville (available online). Tentative syllabus: Class 1(2h): Machine learningbasics. Class 2 (2h): feed forward neural networks, Convolutional neural networks. Class 3 (2h): Recurrent neural networks and representation Class 4(2h): Perception pipelines and main models Class 5 (2h): Edge computing Class 6 (2h): Discussion focused on relevant papers (+2h reading) Lab (5h): introduction to machine learning tools - e.g., PyTorch - (2h), development of basic neural networks (3h)

In presenza

Presentazione orale

P.D.2-2 - Giugno

Where: Clarke meeting room (DET, Cittadella Politecnica, entrance from Corso Castelfidardo 42/A, 1st floor) - Tue. June 20, 13:00-16:00 (lecture) - Thu. June 22, 13:00-16:00 (lecture) - Tue. June 27, 13:00-16:00 (lecture) - Thu. June 29, 13:00-16:00 (lecture) - Tue. July 4, 16:00-18:00 (lab, on Zoom) - Thu. July 6, 15:00-18:00 (lab, on Zoom) - Fri. July 7, 9:00-12:00 (Final exam)