Rule-Based Out-of-Distribution Detection

被引：0
作者：
De Bernardi G. [1 ,2 ,3 ]
Narteni S. [1 ,3 ,4 ]
Cambiaso E. [1 ,3 ]
Mongelli M. [1 ,3 ]
机构：
[1] CNR-Istituto di Elettronica, Ingegneria Dell' Informazione e Delle Telecomunicazioni, Genoa
[2] Universitá Degli Studi di Genova, The Electrical, Electronics and Telecommunication Engineering and Naval Architecture Department, Genova
[3] CNR-Istituto di Elettronica, Ingegneria dell'Informazione e Delle Telecomunicazioni, Genoa
[4] Politecnico di Torino, Department of Control and Computer Engineering, Turin
来源：
IEEE Transactions on Artificial Intelligence | 2024年 / 5卷 / 06期
关键词：
EXplainable AI; mutual information; open data; out-of-distribution detection;
D O I：
10.1109/TAI.2023.3323923
中图分类号：
学科分类号：
摘要：
Out-of-distribution detection is one of the most critical issue in the deployment of machine learning. The data analyst must assure that data in operation should be compliant with the training phase as well as understand if the environment has changed in a way that autonomous decisions would not be safe anymore. The method of the paper is based on eXplainable Artificial Intelligence (XAI); it takes into account different metrics to identify any resemblance between in-distribution and out of, as seen by the XAI model. The approach is nonparametric and distributional assumption free. The validation over complex scenarios (predictive maintenance, vehicle platooning, covert channels in cybersecurity) corroborates both precision in detection and evaluation of training-operation conditions proximity. © 2020 IEEE.
引用
收藏
页码：2627 / 2637
页数：10
相关论文
共 33 条
  • [1] Mirasierra V., Mammarella M., Dabbene F., Alamo T., Prediction error quantification through probabilistic scaling, IEEE Control Syst. Lett., 6, pp. 1118-1123, (2022)
  • [2] Nascita A., Montieri A., Aceto G., Ciuonzo D., Persico V., Pescape A., Improving performance, reliability, and feasibility in multimodal multitask traffic classification with XAI, IEEE Trans. Netw. Serv. Manage., 20, 2, pp. 1267-1289, (2023)
  • [3] Dondio P., Longo L., Trust-based techniques for collective intelligence in social search systems, Next Generation Data Technologies for Collective Computational Intelligence, pp. 113-135, (2011)
  • [4] EASA Concept Paper: First Usable Guidance for Level 1 Machine Learning Applications, (2021)
  • [5] Concepts of Design Assurance for Neural Networks Codann, (2020)
  • [6] Road Vehicles Safety of the Intended Functionality PD ISO PAS 21448:2019, (2019)
  • [7] Heidecker F., Et al., An application-driven conceptualization of corner cases for perception in highly automated driving, Proc. IEEE Intell. Veh. Symp., pp. 644-651, (2021)
  • [8] Cabitza F., Campagner A., The need to separate the wheat from the chaff in medical informatics: Introducing a comprehensive checklist for the (self)-assessment of medical ai studies, Int. J. Med. Inform., 153, (2021)
  • [9] Cabitza F., Campagner A., The need to separate the wheat from the chaff in medical informatics: Introducing a comprehensive checklist for the (self)-assessment of medical AI studies, Int. J. M. Inform., 153, (2021)
  • [10] Sun Y., Ming Y., Zhu X., Li Y., Out-of-distribution detection with deep nearest neighbors, Proc. Int. Conf. Mach. Learn., pp. 20827-20840, (2022)
1234