cover

Decoding Split Window Sensitivity in Signature Isolation Forests

20 Nov 2024

Explore how split windows in Signature Isolation Forests enhance anomaly detection, balancing accuracy and efficiency for isolated and persistent anomalies.

cover

Redefining Anomaly Detection with Signature Isolation Forests

20 Nov 2024

Discover SIF and K-SIF, advanced anomaly detection methods using rough path theory, offering improved accuracy and flexibility over traditional approaches.

cover

Unveiling Path Signatures: A Key to Geometric Insights in Data

20 Nov 2024

Discover the power of path signatures, a sequence of iterated integrals that encapsulate the geometric and topological essence of paths.

cover

How Functional Isolation Forest Detects Anomalies

20 Nov 2024

Explore the mechanics of Functional Isolation Forest (FIF), a cutting-edge anomaly detection method utilizing random splits in functional Hilbert space.

cover

What is the Signature Isolation Forest?

19 Nov 2024

Discover Signature Isolation Forest (SIF), a groundbreaking approach to functional anomaly detection.

cover

Boosting Fairness and Robustness in Over-the-Air Federated Learning: Conclusion and References

27 Oct 2024

This paper presents a federated learning algorithm using Over-the-Air computation for fairness and robustness, optimizing performance in decentralized networks.

cover

Boosting Fairness and Robustness in Over-the-Air Federated Learning: Numerical Example

27 Oct 2024

This paper presents a federated learning algorithm using Over-the-Air computation for fairness and robustness, optimizing performance in decentralized networks.

cover

Boosting Fairness and Robustness in Over-the-Air Federated Learning: Convergence Properties

27 Oct 2024

This paper presents a federated learning algorithm using Over-the-Air computation for fairness and robustness, optimizing performance in decentralized networks.

cover

Boosting Fairness and Robustness in Over-the-Air Federated Learning: FedAir Algorithm

27 Oct 2024

This paper presents a federated learning algorithm using Over-the-Air computation for fairness and robustness, optimizing performance in decentralized networks.