Palestrante

HEITOR SOARES RAMOS FILHO

I am a Computer Science Professor currently working at the Federal University of Minas Gerais, Brazil, with 10+ years of experience in teaching Computer Science classes, advising graduate students, and coordinating R&D projects. My main research interests are focused on data analysis for the Internet of Things, Sensor Networks, Social Networks, and Urban Computing applications. I have been an active member of the scientific community, where I published 70+ scientific papers in different conferences and periodicals and 1 USPTO patent. I have been advising 10+ graduate students and 16+ undergrad projects. 
Specialties: Computer Networks, Sensor Networks, Mobile Computing, Wireless Networks, Data Analysis and Machine Learning

Federated Learning

In this talk, we present the main fundamentals of Federated Learning (FL), covering the steps necessary for the development of applications and services aimed at the Internet of Things (IoT). We introduce the state-of-the-art in FL, challenges, and future perspectives for advancing the field. FL is a recent research field that still has a lot to be explored. Therefore, it has research fronts that remain open, with non-trivial challenges. Thus, we cover the following topics:
  1. The motivation of the area
  2. How FL arises as a possible solution
  3. Privacy-preserving machine learning with FL
  4. The main aspects and characteristics of FL
  5. What are the main LF research problems

Canal YouTube BSIUFRN   Palestra Federated Learning - 25/10/2021 10:00 - 11:00