Technology Category
Machine Learning: Distributed Machine Learning & Federated LearningData Mining & Knowledge Management: Anomaly/Outlier DetectionSearch and Optimization: Learning to Search
Federated Learning with Reservoir State Analysis for Time Series Anomaly Detection
🤖 AI Summary
To address the dual challenges of data non-sharability and low computational efficiency in time-series anomaly detection for privacy-sensitive scenarios, this paper proposes IncFed MD-RS, an incremental federated learning framework. Our method innovatively integrates reservoir computing with Mahalanobis distance–based statistical modeling into the federated pipeline: raw data and gradients remain local, while only lightweight statistics—e.g., means and covariances—are exchanged across clients. This enables gradient-free, online parameter updates. We further introduce subsampling optimization and a heterogeneous sequence–adaptive aggregation mechanism to enhance robustness on short sequences and sparse samples. Evaluated on multiple benchmark datasets, IncFed MD-RS outperforms existing deep and reservoir-based federated approaches in detection accuracy, while significantly reducing communication and computational overhead. The framework achieves a favorable trade-off among privacy preservation, inference efficiency, and generalization capability.
Technology Category
Application Category
📝 Abstract
With a growing data privacy concern, federated learning has emerged as a promising framework to train machine learning models without sharing locally distributed data. In federated learning, local model training by multiple clients and model integration by a server are repeated only through model parameter sharing. Most existing federated learning methods assume training deep learning models, which are often computationally demanding. To deal with this issue, we propose federated learning methods with reservoir state analysis to seek computational efficiency and data privacy protection simultaneously. Specifically, our method relies on Mahalanobis Distance of Reservoir States (MD-RS) method targeting time series anomaly detection, which learns a distribution of reservoir states for normal inputs and detects anomalies based on a deviation from the learned distribution. Iterative updating of statistical parameters in the MD-RS enables incremental federated learning (IncFed MD-RS). We evaluate the performance of IncFed MD-RS using benchmark datasets for time series anomaly detection. The results show that IncFed MD-RS outperforms other federated learning methods with deep learning and reservoir computing models particularly when clients' data are relatively short and heterogeneous. We demonstrate that IncFed MD-RS is robust against reduced sample data compared to other methods. We also show that the computational cost of IncFed MD-RS can be reduced by subsampling from the reservoir states without performance degradation. The proposed method is beneficial especially in anomaly detection applications where computational efficiency, algorithm simplicity, and low communication cost are required.
Problem
Research questions and friction points this paper is trying to address.
Enhance computational efficiency in federated learning
Protect data privacy in time series analysis
Detect anomalies using reservoir state analysis
Innovation
Methods, ideas, or system contributions that make the work stand out.
Federated learning enhances data privacy
Reservoir state analysis detects anomalies
Incremental federated learning reduces computational cost
🔎 Similar Papers
No similar papers found.
