PhD Course: Federated Learning: Challenges, Trends and Emerging Applications

Federated Learning (FL) is a collaborative machine learning technique that trains a model using data from decentralized entities (i.e. IoT devices or Silos). This paradigm has emerging in academic as well as industry since it offer a collaborative way to perform learning task and it preserves data privacy by design. However, when impenitent FL framework in a real world scenario, several challenges come into play (e.g. No—IID, communication costs, computational complexity).

This course introduces students to the fundamentals and recent advances in federated learning. Then, algorithms and methods are presented as means to  handle data heterogeneity, communication and computational complexity.

Table of Contents

• Fundamentals on Federated Learning
• Challenges, Algorithms and Optimization methods
• Federated Learning Frameworks (e.g Flower, Federated TensorFLow)
• Advanced concepts and applications in Federated Learning

• 3CFU
• giorni 13, 15 e 17 aprile ore 9:00-13:00 – Aula Didattica, Cubo 44Z, piano terra