This paper explores using Cluster Validity Indices Fuzzy Adaptative Resonance Theory (CVI Fuzzy ART) to cluster ground motion records (GMRs). Clustering the features extracted from a supervised network trained for predicting the structure damage results in less overfitting from the trained network. Using Cluster Validity Indices (CVIs) to evaluate the clustering gives feedback to how well the data is being classified, allowing further separation of the data. By using CVI Fuzzy ART in combination with features extracted from a trained Convolutional Neural Network (CNN), we were able to form additional clusters in the data. Within the primary clusters, accuracy was improved from our previous best of 82%  up to 95% accuracy. Additionally, the ∼20% of the data that ended up in secondary clusters was identified as borderline data, highlighting that the result is unstable, and that the more severe class should be considered despite the simulation results.
D. Tanksley et al., "Analyzing Ground Motion Records With CVI Fuzzy ART," Proceedings - 2023 IEEE Conference on Artificial Intelligence, CAI 2023, pp. 338 - 340, Institute of Electrical and Electronics Engineers, Jan 2023.
The definitive version is available at https://doi.org/10.1109/CAI54212.2023.00149
Civil, Architectural and Environmental Engineering
Electrical and Computer Engineering
Keywords and Phrases
Adaptive Resonance Theory; Clustering; Earthquakes; Ground Motion Records; Machine Learning; Neural Networks
Article - Conference proceedings
© 2023 Institute of Electrical and Electroncis Engineers, All rights reserved.
01 Jan 2023