Ensemble weighting strategy for federated learning to handle heterogeneous data distributions
By: Richter, Lucas
.
Contributor(s): Dontsov, Ilja.
Publisher: Telangana IUP Publications 2022Edition: Vol.15(4), Oct.Description: 7-20p.Subject(s): Electrical EngineeringOnline resources: Click here
In:
IUP journal of electrical and electronics engineeringSummary: Increasingly measured data in the context of smart cities can be used to develop new and innovative business models to increase efficiency and the value of life. A time-series classification algorithm can support to automatize many different processes such as forecasting services. In order to ensure data security and privacy, Federated Learning trains a global model collaboratively on multiple clients. Having different data-distributions and data-quantities across participating clients, neural networks suffer from slow convergence and overfitting. Based on different data-distributions, data-quantities and number of clients, we develop and evaluate different data-clustering strategies to update global model weights in comparison to the state of the art. We use public time-series data, generate various synthetic datasets and train a Relational-Regularized Autoencoder for classification purposes. Our results show an improvement of model performance concerning generalization.
| Item type | Current location | Call number | Status | Date due | Barcode | Item holds |
|---|---|---|---|---|---|---|
Articles Abstract Database
|
School of Engineering & Technology Archieval Section | Not for loan | 2023-0627 |
Increasingly measured data in the context of smart cities can be used to develop new and innovative business models to increase efficiency and the value of life. A time-series classification algorithm can support to automatize many different processes such as forecasting services. In order to ensure data security and privacy, Federated Learning trains a global model collaboratively on multiple clients. Having different data-distributions and data-quantities across participating clients, neural networks suffer from slow convergence and overfitting. Based on different data-distributions, data-quantities and number of clients, we develop and evaluate different data-clustering strategies to update global model weights in comparison to the state of the art. We use public time-series data, generate various synthetic datasets and train a Relational-Regularized Autoencoder for classification purposes. Our results show an improvement of model performance concerning generalization.
Click on an image to view it in the image viewer
There are no comments for this item.