On the performance of deep learning models for time series classification in streaming
This work addresses the problem of real-time data streaming classification for applications requiring fast predictions, but it is incremental as it adapts existing deep learning methods to a streaming context.
The study evaluated deep learning models for time series classification in streaming scenarios using an asynchronous dual-pipeline framework, finding that convolutional architectures achieved higher accuracy and efficiency compared to other models like MLPs and RNNs.
Processing data streams arriving at high speed requires the development of models that can provide fast and accurate predictions. Although deep neural networks are the state-of-the-art for many machine learning tasks, their performance in real-time data streaming scenarios is a research area that has not yet been fully addressed. Nevertheless, there have been recent efforts to adapt complex deep learning models for streaming tasks by reducing their processing rate. The design of the asynchronous dual-pipeline deep learning framework allows to predict over incoming instances and update the model simultaneously using two separate layers. The aim of this work is to assess the performance of different types of deep architectures for data streaming classification using this framework. We evaluate models such as multi-layer perceptrons, recurrent, convolutional and temporal convolutional neural networks over several time-series datasets that are simulated as streams. The obtained results indicate that convolutional architectures achieve a higher performance in terms of accuracy and efficiency.