Explainable feature selection by SHAP for signal recognition of Ф-OTDR system

Min Sun; Nian Fang

doi:10.1117/12.3007857

24 November 2023 Explainable feature selection by SHAP for signal recognition of Φ-OTDR system

Min Sun, Nian Fang

Proceedings Volume 12935, Fourteenth International Conference on Information Optics and Photonics (CIOP 2023); 1293540 (2023) https://doi.org/10.1117/12.3007857
Event: Fourteenth International Conference on Information Optics and Photonics (CIOP 2023), 2023, Xi’an, China

Abstract

An explainable feature selection method based on Shapley additive explanation (SHAP) is proposed for the signal recognition of a phase-sensitive optical time-domain reflectometer system. The SHAP value is used to quantify the contribution of a feature. The original features of the signals to be identified are selected according their contributions. The support vector machine algorithm is employed as the classifier to verify the effectiveness and explanation of the proposed feature selection method in the signal recognition on an open dataset from Beijing Jiaotong University. The average recognition accuracy of six types of signals increases to 89.3% and the corresponding recognition time of each sample reduces to 0.097 s. The feature selection method provides a reliable guidance for feature selection and can improve the speed and accuracy of the signal recognition.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Min Sun and Nian Fang "Explainable feature selection by SHAP for signal recognition of Φ-OTDR system", Proc. SPIE 12935, Fourteenth International Conference on Information Optics and Photonics (CIOP 2023), 1293540 (24 November 2023); https://doi.org/10.1117/12.3007857

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available