Sem: Sept-Dec 2024

Green Federated Learning

Pre-requisite: Knowledge of programming, and Machine Learning.
Syllabus:
  • Introduction: Federated Learning Vs. Green Federated Learning
    • What is Federated Learning?, Impact of GFL, GFL Challenges, GFL Applications
    • Federated Learning vs Green Federated Learning
    • Green Federated Learning Applications
  • Traditional federated Learning Algorithms
    • Federated Averaging (FedAvg)
    • Federated Proximity (FedProx)
  • Challenges in Green Federated Learning
    • Convergence
    • Communication Rounds
    • Carbon Emmission
    • Accuracy
  • Algorithms and methods in Green FL
    • Aggregation
    • Adaptive
  • Methods to address the challenges of green federated learning
    • Client Selection
    • Quantization
    • Model Pruning
  • Optimization methods in Green federated learning
    • SGD
    • Adaptive SGD
    • Adaptive Learning Rate
    Text Book:

    Reference Books & Resources:

    Instructors

    • Dr. Dipanwita Thakur

    Class Timing and Venue

    • Class Timing:
    • Venue: TBD
    • Mode of Lectures: Live Class

    Grading Strategy

    • Class Participation Quiz: 10%
    • Mid-Sem Exam: 25%
    • End-Sem Exam: 35%
    • GFL Project: 15%
    • Research Paper Presentation: 15%

    Classes

    Sl.No.Lecture DateTopics Taught in ClassDetailed ResourcesNotes/PPT
    1.16th August'24Course Plan----

    Announcement