Current browse context:
cs.LG
Change to browse by:
References & Citations
Computer Science > Machine Learning
Title: Input Convex Lipschitz RNN: A Fast and Robust Approach for Engineering Tasks
(Submitted on 15 Jan 2024 (v1), revised 19 Jan 2024 (this version, v2), latest version 27 Mar 2024 (v3))
Abstract: Computational efficiency and adversarial robustness are critical factors in real-world engineering applications. Yet, conventional neural networks often fall short in addressing both simultaneously, or even separately. Drawing insights from natural physical systems and existing literature, it is known that an input convex architecture enhances computational efficiency, while a Lipschitz-constrained architecture bolsters adversarial robustness. By leveraging the strengths of convexity and Lipschitz continuity, we develop a novel network architecture, termed Input Convex Lipschitz Recurrent Neural Networks. This model outperforms existing recurrent units across a spectrum of engineering tasks in terms of computational efficiency and adversarial robustness. These tasks encompass a benchmark MNIST image classification, real-world solar irradiance prediction for Solar PV system planning at LHT Holdings in Singapore, and real-time Model Predictive Control optimization for a chemical reactor.
Submission history
From: Zihao Wang [view email][v1] Mon, 15 Jan 2024 06:26:53 GMT (11427kb,D)
[v2] Fri, 19 Jan 2024 06:16:59 GMT (11427kb,D)
[v3] Wed, 27 Mar 2024 16:06:34 GMT (12286kb,D)
Link back to: arXiv, form interface, contact.