Federated Reconstruction for Matrix Factorization (Building recommendation systems with TensorFlow)
Looking to train models for on-device inference without gathering any sensitive user data? Developer Advocate Wei Wei talks about Federated Reconstruction for matrix factorization, a novel technique for building recommendation systems using TensorFlow Federated (TFF). Follow along as he takes you through a cross-device federated learning example.
Resources:
Federated learning video→ https://goo.gle/3qttKIM
TensorFlow Federated → https://goo.gle/3twlycG
Collaborative learning video → https://goo.gle/37Wd0DB
Federated Reconstruction for Matrix Factorization → https://goo.gle/3wwBRYP
A Scalable Approach for Partially Local Federated Learning → https://goo.gle/3wukl7o
Federated Reconstruction for Matrix Factorization tutorial → https://goo.gle/3wwBRYP
Federated Reconstruction: Partially Local Federated Learning paper → https://goo.gle/3isZNnx
TFF FedRecon libraries → https://goo.gle/3wwhLxG
Federated Learning Workshop - FLA Research Demos & TFF Tutorials → https://goo.gle/3D3i2cZ
Chapters:
0:00 - Introduction
0:55 - What is federated learning?
1:40 - Cross-device federated learning example
5:42 - Code walkthrough
7:15 - Wrap up
Watch more Building recommendation systems with TensorFlow → https://goo.gle/3Bi8NUS
Subscribe to TensorFlow → https://goo.gle/TensorFlow
product: TensorFlow - TensorFlow Recommenders; fullname: Wei Wei;