<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>Predicting human mobility between locations plays an important role in a wide range of applications and services such as transportation, economics, sociology and other fields. Mobility prediction can be implemented through various machine learning algorithms that can predict the future trajectory of a user relying on the current trajectory and time, learning from historical sequences of locations previously visited by the user. But, it is not easy to capture complex patterns from the long historical sequences of locations. Inspired by the methods of the Convolutional Neural Network (CNN), we propose an augmented Union ConvAttention-LSTM (UCAL) model. The UCAL consists of the 1D CNN that allows capturing locations from historical trajectories and the augmented proposed model that contains an Attention technique with a Long Short-Term Memory (LSTM) in order to capture patterns from current trajectories. The experimental results prove the effectiveness of our proposed methodology that outperforms the existing models.</p>
</abstract>

Predicting human mobility between locations plays an important role in a wide range of applications and services such as transportation, economics, sociology and other fields. Mobility prediction can be implemented through various machine learning algorithms that can predict the future trajectory of a user relying on the current trajectory and time, learning from historical sequences of locations previously visited by the user. But, it is not easy to capture complex patterns from the long historical sequences of locations. Inspired by the methods of the Convolutional Neural Network (CNN), we propose an augmented Union ConvAttention-LSTM (UCAL) model. The UCAL consists of the 1D CNN that allows capturing locations from historical trajectories and the augmented proposed model that contains an Attention technique with a Long Short-Term Memory (LSTM) in order to capture patterns from current trajectories. The experimental results prove the effectiveness of our proposed methodology that outperforms the existing models.

An Augmented UCAL Model for Predicting Trajectory and Location

Cybernetics and Information Technologies

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

{"article-title":"An Augmented UCAL Model for Predicting Trajectory and Location"}

Predicting human mobility between locations plays an important role in a wide range of applications and services such as transportation, economics, sociology...

An Augmented UCAL Model for Predicting Trajectory and Location

Online veröffentlicht: 23. Juni 2022

Seitenbereich: 114 - 124

Eingereicht: 24. Jan. 2022

Akzeptiert: 29. März 2022

DOI: https://doi.org/10.2478/cait-2022-0020

Schlüsselwörter
Deep learning, LSTM, attention mechanism, human mobility prediction location, trajectory

© 2022 Nesrine Kadri et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

An Augmented UCAL Model for Predicting Trajectory and Location

Online veröffentlicht: 23. Juni 2022

Seitenbereich: 114 - 124

Eingereicht: 24. Jan. 2022

Akzeptiert: 29. März 2022

DOI: https://doi.org/10.2478/cait-2022-0020

SchlüsselwörterDeep learning, LSTM, attention mechanism, human mobility prediction location, trajectory

© 2022 Nesrine Kadri et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Schlüsselwörter
Deep learning, LSTM, attention mechanism, human mobility prediction location, trajectory