Distance-based one-class time-series classification approach using local cluster balance

被引：0

作者：

Hayashi, Toshitaka ^{[1
]}

Cimr, Dalibor ^{[1
]}

Studnička, Filip ^{[1
]}

Fujita, Hamido ^{[2
,3
,4
]}

Bušovský, Damián ^{[1
]}

Cimler, Richard ^{[1
]}

Selamat, Ali ^{[2
]}

机构：

[1] Faculty of Science, University of Hradec Kralove, Hradecká 1285, Hradec Kralove, Czech Republic

[2] Malaysia-Japan International Institute of Technology (MJIIT), Universiti Teknologi Malaysia, Kuala Lumpur,54100, Malaysia

[3] Andalusian Research Institute in Data Science and Computational Intelligence (DaSCI), University of Granada, Granada, Spain

[4] Regional Research Center, Iwate Prefectural University, Takizawa,020-0693, Japan

来源：

Expert Systems with Applications | 2024年 / 235卷

基金：

日本学术振兴会;

关键词：

Communication channels (information theory) - Decoding - Time series;

D O I：

暂无

中图分类号：

学科分类号：

摘要：

Deciding the signal length is an important challenge for one-class time-series classification (OCTSC). This paper aims to develop an OCTSC algorithm that does not require model retraining for different signal lengths. For this purpose, a distance-based one-class time-series classification approach using local cluster balance (OCLCB) is proposed. OCLCB extracts feature vectors, namely, local cluster balance (LCB), from the clustering results of sliding windows. K-means clustering is applied to the sliding windows extracted from the training signal. Then, the local prototype (LP) is calculated as the average of the local cluster balance (LCB) in the training data. Unseen scores are computed as the distance metrics between LP and LCBs in the testing data. Since the sliding window size is independent of the entire signal size, OCLCB does not need to retrain the model. This aspect gives the benefit of reducing the parameter tuning costs. The source code is uploaded at https://github.com/ToshiHayashi/OCLCB. © 2023 Elsevier Ltd

引用