Subwork-template

The objectives & results should be detailed, including time, what to do, how to do it, and problems encountered.

Expected Timeline: --; Contributor: --

Objectives

What to do. To achieve these objectives, what should you learn, e.g., techniques? For example, to start research on federated learning:

• Challenges in FL

  • Statistical Heterogeneity

  • System Heterogeneity

• Existing Solutions

  • Algorithms: FedAvg, FedProx...

• Improve FedAvg on Non-IID

  • Algorithm implementation (PyTorch or Tensorflow)

  • Theorem for bounded weight divergence

  • Experiment setup

Process

How to do. To achieve your objectives, how do you make progress, e.g., reading papers? For example, to start research on federated learning:

Objectives	Process	Time
Challenges in FL	Read survey papers (...)	2023.1.1-1.10
Existing Solutions	Read corresponding papers	2023.1.11-1.13
Improve FedAvg on Non-IID-Stage1	Implement FedAvg using PyTorch (simulation)	2023.1.15-1.20

Reflections & Thoughts (Optional)

What you learn. After each process you made, what have you gained and your innovative insights? For example, to start research on federated learning:

• Challenges in FL from survey papers: the lack of information for optimization, the tradeoff between statistical and system utility, and expensive Communication (ref. ...)

Problems Encountered & Solutions

What problems have you encountered, and how did you solve them?

Conclusion