Redirigiendo al acceso original de articulo en 20 segundos...
Inicio  /  Applied Sciences  /  Vol: 12 Par: 7 (2022)  /  Artículo
ARTÍCULO
TITULO

Self-Attentive Moving Average for Time Series Prediction

Yaxi Su    
Chaoran Cui and Hao Qu    

Resumen

Time series prediction has been studied for decades due to its potential in a wide range of applications. As one of the most popular technical indicators, moving average summarizes the overall changing patterns over a past period and is frequently used to predict the future trend of time series. However, traditional moving average indicators are calculated by averaging the time series data with equal or predefined weights, and ignore the subtle difference in the importance of different time steps. Moreover, unchanged data weights will be applied across different time series, regardless of the differences in their inherent characteristics. In addition, the interaction between different dimensions of different indicators is ignored when using the moving averages of different scales to predict future trends. In this paper, we propose a learning-based moving average indicator, called the self-attentive moving average (SAMA). After encoding the input signals of time series based on recurrent neural networks, we introduce the self-attention mechanism to adaptively determine the data weights at different time steps for calculating the moving average. Furthermore, we use multiple self-attention heads to model the SAMA indicators of different scales, and finally combine them through a bilinear fusion network for time series prediction. Extensive experiments on two real-world datasets demonstrate the effectiveness of our approach. The data and codes of our work have been released.

 Artículos similares

       
 
Huiying Ren, Z. Jason Hou, Mark Wigmosta, Ying Liu and L. Ruby Leung    
Changes in extreme precipitation events may require revisions of civil engineering standards to prevent water infrastructures from performing below the designated guidelines. Climate change may invalidate the intensity-duration-frequency (IDF) computatio... ver más
Revista: Water

 
Francisco Leitão, Vânia Baptista, Vasco Vieira, Patrícia Laginha Silva, Paulo Relvas and Maria Alexandra Teodósio    
Coastal upwelling has a significant local impact on marine coastal environment and on marine biology, namely fisheries. This study aims to evaluate climate and environmental changes in upwelling trends between 1950 and 2010. Annual, seasonal and monthly ... ver más
Revista: Water

 
Igor Paz, Bernard Willinger, Auguste Gires, Bianca Alves de Souza, Laurent Monier, Hervé Cardinal, Bruno Tisserand, Ioulia Tchiguirinskaia and Daniel Schertzer    
Recent studies have highlighted the need for high resolution rainfall measurements for better modelling of urban and peri-urban catchment responses. In this work, we used a fully-distributed model called ?Multi-Hydro? to study small-scale rainfall variab... ver más
Revista: Water

 
Kenan Liu, Wuyun Zhao, Bugong Sun, Pute Wu, Delan Zhu and Peng Zhang    
Autonomous navigation for agricultural machinery has broad and promising development prospects. Kalman filter technology, which can improve positioning accuracy, is widely used in navigation systems in different fields. However, there has not been much r... ver más
Revista: Water

 
Zhenzhen Di, Miao Chang, Peikun Guo, Yang Li and Yin Chang    
Most worldwide industrial wastewater, including in China, is still directly discharged to aquatic environments without adequate treatment. Because of a lack of data and few methods, the relationships between pollutants discharged in wastewater and those ... ver más
Revista: Water