VGU RESEARCH REPOSITORY
Please use this identifier to cite or link to this item:
https://epub.vgu.edu.vn/handle/dlibvgu/1862| Title: | Local temporal feature for timeseries prediction | Authors: | Tran Ngoc Duy Chuong | Keywords: | Air quality data management;Time series data | Issue Date: | 2024 | Abstract: | Gaps in time series data pose significant challenges for deep learning research, impacting the performance of predictive models. This thesis proposes a novel hybrid model to address these challenges by combining one-dimensional convolutional layers (Conv1D), inspired by the Inception architecture, with Long Short-Term Memory (LSTM) units. This innovative model leverages the spatial feature extraction capabilities of Conv1D and the temporal sequence learning strengths of LSTM to improve prediction accuracy.Our model was rigorously tested on datasets characterized by extensive gaps and dense data points, specifically focusing on air quality measurements at Vietnamese German University. The results demonstrate the model’s superior ability to handle low-quality time series data without the need for manual adjustments specific to each dataset, highlighting its robustness and versatility |
URI(1): | https://epub.vgu.edu.vn/handle/dlibvgu/1862 | Rights: | Attribution-NonCommercial 4.0 International |
| Appears in Collections: | Computer Science (CS) |
Files in This Item:
| File | Description | Size | Format | Existing users please Login |
|---|---|---|---|---|
| Local temporal feature for timeseries prediction.pdf | 15.87 MB | Adobe PDF |
Page view(s)
45
checked on Aug 24, 2025
Download(s)
18
checked on Aug 24, 2025
Google ScholarTM
Check
This item is licensed under a Creative Commons License