Certifiably Robust Neural ODE with Learning-based Barrier Function. Yang, R., Jia, R., Zhang, X., & Jin, M. IEEE Control Systems Letters (Special Issue on Data-Driven Analysis and Control), 2023.
![Certifiably Robust Neural ODE with Learning-based Barrier Function [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Pdf abstract bibtex 19 downloads
Pdf abstract bibtex 19 downloads Neural Ordinary Differential Equations (ODEs) have gained traction in many applications. While recent studies have focused on empirically increasing the robustness of neural ODEs against natural or adversarial attacks, certified robustness is still lacking. In this work, we propose a framework for training a neural ODE using barrier functions and demonstrate improved robustness for classification problems. We further provide the first generalization guarantee of robustness against adversarial attacks using a wait-and-judge scenario approach.
@article{2023_3C_BNODE,
title={Certifiably Robust Neural ODE with Learning-based Barrier Function},
author={Runing Yang and Ruoxi Jia and Xiangyu Zhang and Ming Jin},
journal={IEEE Control Systems Letters (Special Issue on Data-Driven Analysis and Control)},
pages={},
year={2023},
url_pdf={B-NODE22.pdf},
keywords = {Optimization, Control theory, Machine Learning},
abstract={Neural Ordinary Differential Equations (ODEs) have gained traction in many applications. While recent studies have focused on empirically increasing the robustness of neural ODEs against natural or adversarial attacks, certified robustness is still lacking. In this work, we propose a framework for training a neural ODE using barrier functions and demonstrate improved robustness for classification problems. We further provide the first generalization guarantee of robustness against adversarial attacks using a wait-and-judge scenario approach. },
}Downloads: 19
{"_id":"J3BJTKqyfvLGQDfRa","bibbaseid":"yang-jia-zhang-jin-certifiablyrobustneuralodewithlearningbasedbarrierfunction-2023","author_short":["Yang, R.","Jia, R.","Zhang, X.","Jin, M."],"bibdata":{"bibtype":"article","type":"article","title":"Certifiably Robust Neural ODE with Learning-based Barrier Function","author":[{"firstnames":["Runing"],"propositions":[],"lastnames":["Yang"],"suffixes":[]},{"firstnames":["Ruoxi"],"propositions":[],"lastnames":["Jia"],"suffixes":[]},{"firstnames":["Xiangyu"],"propositions":[],"lastnames":["Zhang"],"suffixes":[]},{"firstnames":["Ming"],"propositions":[],"lastnames":["Jin"],"suffixes":[]}],"journal":"IEEE Control Systems Letters (Special Issue on Data-Driven Analysis and Control)","pages":"","year":"2023","url_pdf":"B-NODE22.pdf","keywords":"Optimization, Control theory, Machine Learning","abstract":"Neural Ordinary Differential Equations (ODEs) have gained traction in many applications. While recent studies have focused on empirically increasing the robustness of neural ODEs against natural or adversarial attacks, certified robustness is still lacking. In this work, we propose a framework for training a neural ODE using barrier functions and demonstrate improved robustness for classification problems. We further provide the first generalization guarantee of robustness against adversarial attacks using a wait-and-judge scenario approach. ","bibtex":"@article{2023_3C_BNODE,\n title={Certifiably Robust Neural ODE with Learning-based Barrier Function},\n author={Runing Yang and Ruoxi Jia and Xiangyu Zhang and Ming Jin},\n journal={IEEE Control Systems Letters (Special Issue on Data-Driven Analysis and Control)},\n pages={},\n year={2023},\n url_pdf={B-NODE22.pdf},\n keywords = {Optimization, Control theory, Machine Learning},\n abstract={Neural Ordinary Differential Equations (ODEs) have gained traction in many applications. While recent studies have focused on empirically increasing the robustness of neural ODEs against natural or adversarial attacks, certified robustness is still lacking. In this work, we propose a framework for training a neural ODE using barrier functions and demonstrate improved robustness for classification problems. We further provide the first generalization guarantee of robustness against adversarial attacks using a wait-and-judge scenario approach. },\n}\n\n","author_short":["Yang, R.","Jia, R.","Zhang, X.","Jin, M."],"key":"2023_3C_BNODE","id":"2023_3C_BNODE","bibbaseid":"yang-jia-zhang-jin-certifiablyrobustneuralodewithlearningbasedbarrierfunction-2023","role":"author","urls":{" pdf":"http://www.jinming.tech/papers/B-NODE22.pdf"},"keyword":["Optimization","Control theory","Machine Learning"],"metadata":{"authorlinks":{}},"downloads":19},"bibtype":"article","biburl":"http://www.jinming.tech/papers/myref.bib","dataSources":["sTzDHHaipTZWjp8oe","Y64tp2HnDCfXgLdc5"],"keywords":["optimization","control theory","machine learning"],"search_terms":["certifiably","robust","neural","ode","learning","based","barrier","function","yang","jia","zhang","jin"],"title":"Certifiably Robust Neural ODE with Learning-based Barrier Function","year":2023,"downloads":19}