Scalable auto-encoders for gravitational waves detection from time series data
作者：
Highlights：
• Deep learning approaches to analyze time series from Gravitational Waves detectors.
• Autoencoders are trained with one-class data for feature extraction and classification.
• The non-linear extracted features improve the Gravitational Waves detection task.
• High accuracy results on three datasets for Gravitational Waves detection.
• High scalability results with large-scale data on a Spark cluster environment.
摘要
•Deep learning approaches to analyze time series from Gravitational Waves detectors.•Autoencoders are trained with one-class data for feature extraction and classification.•The non-linear extracted features improve the Gravitational Waves detection task.•High accuracy results on three datasets for Gravitational Waves detection.•High scalability results with large-scale data on a Spark cluster environment.
论文关键词：Time series classification,Anomaly detection,Feature extraction,Deep neural networks,Machine learning,Big data analytics,Apache spark,Hadoop
论文评审过程：Received 23 October 2019, Revised 5 March 2020, Accepted 9 March 2020, Available online 14 March 2020, Version of Record 4 April 2020.
论文官网地址：https://doi.org/10.1016/j.eswa.2020.113378
