Apr 21, 2020 · This paper proposes a novel unsupervised pre-training method adapted for FL, which aims to reduce both the communication and computation costs through model ...

This paper proposes a novel unsupervised pre-training method adapted for FL, which aims to reduce both the communication and computation costs through model ...

This decentralized training approach is a practical solution to mitigate the risk of privacy leakage in Deep Neural Network (DNN) model training, as only the ...

Federated learning (FL) enables a neural network (NN) to be trained using privacy-sensitive data on mobile devices while retaining all the data on their ...

Lottery Hypothesis based Unsupervised Pre-training for Model Compression in Federated Learning · Computer Science. IEEE Vehicular Technology Conference · 2020.

The key idea of the proposed method is to obtain a ``good'' subnetwork from the original NN using the unlabeled data based on the lottery hypothesis. Denoising ...

2022. Lottery Hypothesis based Unsupervised Pre-training for Model Compression in Federated Learning. S Itahara, T Nishio, M Morikura, K Yamamoto. IEEE VTC ...

Mar 12, 2024 · The Lottery Ticket Hypothesis (LTH) states that a dense neural network model contains a highly sparse subnetwork (i.e., winning tickets) that ...

2022. Lottery Hypothesis based Unsupervised Pre-training for Model Compression in Federated Learning. S Itahara, T Nishio, M Morikura, K Yamamoto. IEEE VTC ...

The key idea of the proposed method is to obtain a ``good'' subnetwork from the original NN using the unlabeled data based on the lottery hypothesis.