End-to-end multivariate time series classification via hybrid deep learning architectures
- PDF / 1,520,705 Bytes
- 15 Pages / 595.276 x 790.866 pts Page_size
- 95 Downloads / 228 Views
ORIGINAL ARTICLE
End-to-end multivariate time series classification via hybrid deep learning architectures Mehak Khan 1
&
Hongzhi Wang 1 & Alladoumbaye Ngueilbaye 1 & Aya Elfatyany 1
Received: 30 October 2019 / Accepted: 22 August 2020 # Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract Deep learning has revolutionized many areas, including time series data mining. Multivariate time series classification (MTSC) remained to be a well-known problem in the time series data mining community, due to its availability in various practical applications such as healthcare, finance, geoscience, and bioinformatics. Recently, multivariate long short-term memory with fully convolutional network (MLSTM-FCN) and multivariate attention long short-term memory with fully convolutional network (MALSTM-FCN) have shown superior results over various state-of-the-art methods. So, in this paper, we explore the usage of recurrent neural network (RNN), and its variants, such as bidirectional recurrent neural network (BiRNN), bidirectional long short-term memory (BiLSTM), gated recurrent unit (GRU), and bidirectional gated recurrent unit (BiGRU). We augment these RNN variants separately by replacing long short-term memory (LSTM) in MLSTM-FCN, which is the combination of LSTM, squeeze-and-excitation (SE) block, and fully convolutional network (FCN). Moreover, we integrate the SE block within FCN to leverage its high performance for the MTSC task. The resulting algorithms do not require heavy pre-processing or feature crafting. Thus, they could be easily deployed on real-time systems. We conduct a comprehensive evaluation with a large number of standard datasets and demonstrate that our approaches achieve notable results over the current best MTSC approach. Keywords Convolutional neural network . Squeeze-and-excitation . Recurrent neural networks . Long short-term memory . Gated recurrent unit . Multivariate time series classification
1 Introduction Time series classification (TSC) has received much attentiveness in data mining, in which the goal is to classify the data points over time based on its behavior [1]. Time series data is ubiquitous and used in statistics, finance, weather forecasting, pattern recognition, econometrics, astronomy, earthquake
* Mehak Khan [email protected] Hongzhi Wang [email protected] Alladoumbaye Ngueilbaye [email protected] Aya Elfatyany [email protected] 1
School of Computer Science and Technology, Harbin Institute of Technology, Harbin 150001, China
prediction, signal processing, and others. In short, practically any field which includes temporal measurements [2]. A time series dataset could be univariate, in which a single observation recorded sequentially over the equal time interval, whereas multivariate, where multiple time series observations are available simultaneously. The complexity of MTSC is increased due to the data type. The MTSC data comprises of interactions between multiple values at the single timestamp. The problem of MTSC is an open challenge. To solve
Data Loading...
