2026-10-11T16:33:30Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33813718 2026-09-15T17:52:38Z category_4 category_734 category_931 portal_5 item_type_3 month_year_09_2026

<p>Bayesian threshold optimization process for FastDTW.</p> Le Liu (167133) Hongsheng Xu (24949841) Hai Tang (5480003) Jianing Cui (16384754) Li Feng (77923) Ke Deng (278265) Biochemistry Biological Sciences not elsewhere classified Information Systems not elsewhere classified thus markedly improving term memory networks several critical limitations cosine similarity yields optimal detection threshold low detection accuracy online battery diagnosis labeled fault data experimental results show driving conditions based cosine </ p detection performance changes fixed model structure battery anomaly detection reconstruction error metrics anomaly detection experimental conditions battery data detection performance reconstruction mechanism battery datasets best performance trained separately temporal modeling systematically evaluated study offers practical solution paper presents model performance long short highly dependent corresponding subsets commercial vehicles bayesian optimization autoencoder architecture acid batteries <p>Bayesian threshold optimization process for FastDTW.</p> 2026-09-15T17:52:20Z Dataset Dataset 10.1371/journal.pone.0357237.t007 https://figshare.com/articles/dataset/_p_Bayesian_threshold_optimization_process_for_FastDTW_p_/33813718 CC BY 4.0