MPEG-G Microbiome Classification Challenge
$5 000 USD
Completed (~1 year ago)
Classification
Federated Learning
Python
Deep Learning
802 joined
83 active
Start
20 Jun 25
Close
15 Sep 25
Reveal
15 Sep 25
'/%3e%3cpath fill='%23fff' d='m8.751-17.959 10.11 11.373L3.997-9.844l13.94-6.1-7.692 13.129z'/%3e%3c/svg%3e)
Federated learning ................?
Notebooks · 13 Sep 2025, 20:09 · 17
Hello everyone,
To those using federated learning for this competition: would you be willing to share your public LB score and how it compares to your local CV?
My federated learning setup currently achieves a public LB score of 0.005. However, I'm aware that the way I'm splitting the data—randomly across three clients—might be problematic.
by the way i get 0.07 Four clients
im using nn
I am using deep learning
are you using flower ?
yes i am using
how do you defind clients ?
I am working with some teameate my Team mate decide the criteria
How are your federated scores in the private @__yassine__ & @MuhammadQasimShabbeer?
0.026250816
That's a really strong score. Good job
Thanks a lot! I really appreciate it 🙏.
What is yours?
Its 0.02xx. But was tough fixing bugs when I started. So any 0.0xxx is a good score
Congrats on the great score ! That’s really impressive. By the way, did you use Flower or PySyft for the federated setup?
Did not try PySyft looked tough been around for longer than Flower
Ah, got it! I used Flower for my federated setup — found it easier to get started with