BiLSTM-MLAM: A Multi-Scale Time Series Prediction Model for Sensor Data Based on Bi-LSTM and Local Attention Mechanisms

BiLSTM-MLAM: A Multi-Scale Time Series Prediction Model for Sensor Data Based on Bi-LSTM and Local Attention Mechanisms

This paper introduces BiLSTM-MLAM, a novel multi-scale time series prediction model. Initially, the approach utilizes bidirectional long short-term memory to capture information from both forward and backward directions in time series data. Subsequently, a multi-scale patch segmentation module gener...

Full description

Bibliographic Details
Main Authors: Yongxin Fan, Qian Tang, Yangming Guo, Yifei Wei
Format: Article
Language:English
Published: MDPI AG 2024-06-01
Series:Sensors
Subjects:
bidirectional long short-term memory
local attention mechanism
multi-scale patch segmentation
time series prediction
Online Access:https://www.mdpi.com/1424-8220/24/12/3962

Internet

https://www.mdpi.com/1424-8220/24/12/3962

Similar Items