Timezone: »

Architecture Agnostic Federated Learning for Neural Networks
Disha Makhija · Xing Han · Nhat Ho · Joydeep Ghosh

Wed Jul 20 01:50 PM -- 01:55 PM (PDT) @ Room 318 - 320

With growing concerns regarding data privacy and rapid increase in data volume, Federated Learning (FL) has become an important learning paradigm. However, jointly learning a deep neural network model in a FL setting proves to be a non-trivial task because of the complexities associated with the neural networks, such as varied architectures across clients, permutation invariance of the neurons, and presence of non-linear transformations in each layer. This work introduces a novel framework, Federated Heterogeneous Neural Networks (FedHeNN), that allows each client to build a personalised model without enforcing a common architecture across clients. This allows each client to optimize with respect to local data and compute constraints, while still benefiting from the learnings of other (potentially more powerful) clients. The key idea of FedHeNN is to use the instance-level representations obtained from peer clients to guide the simultaneous training on each client. The extensive experimental results demonstrate that the FedHeNN framework is capable of learning better performing models on clients in both the settings of homogeneous and heterogeneous architectures across clients.

Author Information

Disha Makhija (University of Texas at Austin)
Xing Han (The University of Texas at Austin)
Nhat Ho (University of Texas at Austin)
Joydeep Ghosh (The University of Texas at Austin)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors